direct selection for mutators in escherichia coli · low spontaneous mutation rate might not...

JOURNAL OF BACTERIOLOGY,0021-9193/99/$04.0010

Mar. 1999, p. 1576–1584 Vol. 181, No. 5

Copyright © 1999, American Society for Microbiology. All Rights Reserved.

Direct Selection for Mutators in Escherichia coliJEFFREY H. MILLER,* ANJALI SUTHAR, JENNIFER TAI, ANNIE YEUNG,

CINDY TRUONG, AND JEAN LEE STEWART

Department of Microbiology and Molecular Genetics and The Molecular BiologyInstitute, University of California, Los Angeles, California 90095

Received 28 August 1998/Accepted 8 December 1998

We have constructed strains that allow a direct selection for mutators of Escherichia coli on a single platemedium. The plate selection is based on using two different markers whose reversion is enhanced by a givenmutator. Plates containing limiting amounts of each respective nutrient allow the growth of ghost colonies ormicrocolonies that give rise to full-size colonies only if a reversion event occurs. Because two successivemutational events are required, mutator cells are favored to generate full-size colonies. Reversion of a thirdmarker allows direct visualization of the mutator phenotype by the large number of blue papillae in the full-sizecolonies. We also describe plate selections involving three successive nutrient markers followed by a fourthpapillation step. Different frameshift or base substitution mutations are used to select for mismatch-repair-defective strains (mutHLS and uvrD). We can detect and monitor mutator cells arising spontaneously, atfrequencies lower than 1025 in the population. Also, we can measure a mutator cascade, in which one type ofmutator (mutT) generates a second mutator (mutHLS) that then allows stepwise frameshift mutations. Wediscuss the relevance of mutators arising on a single medium as a result of cells overcoming successive growthbarriers to the development and progression of cancerous tumors, some of which are mutator cell lines.

Several successive genetic alterations are required for a can-cer cell to develop, proliferate, and be able to metastasize (70).In fact, as many as 6 to 10 mutations may be needed to turn anormal cell into a full-blown invasive tumor line (2), which isone reason why many cancers take decades to develop. Loeb(35) has postulated that an early step in the progression ofsome cancers may be the creation of a mutator cell, since thelow spontaneous mutation rate might not account for the in-cidence of cancer. However, the elevated mutation rates inmutators would greatly increase the probability of cells accu-mulating all of the necessary mutations (27, 35), especially ifrepetitive rounds of clonal expansion and somatic selectionoccurred (48). In accord with this idea was the exciting findingthat the inherited susceptibility to human nonpolyposis coloncancer (HNPCC) and ovarian cancer is due to a defect in onecopy of one of the genes involved in the human counterpart tothe bacterial mismatch repair system (3, 18, 30, 49). Presum-ably, when a somatic cell loses the other copy, the resulting cellis completely defective for mismatch repair and has a highermutation rate. In fact, tumor lines from HNPCC patients aremutators with greatly increased repeat-tract or microsatelliteinstability (1, 26, 30, 36, 52, 53, 69), a propensity for frequentadditions or deletions at repetitive nucleotide sequences (suchas repetitive mono-, di-, tri-, and tetranucleotide repeats), inanalogy with the repeat-tract instability seen in mismatch-re-pair-deficient strains of bacteria and yeast (11, 57, 66). Thesetumor lines also show elevated mutation rates in genes such ashprt (1, 3, 21). The realization that mutator cells are cancerprone (1, 30, 50, 51) has led to increased research in this area,including a search for new types of mutators and for a betterunderstanding of how mutators proliferate.

Experiments with bacteria aimed at finding mutators (12, 23,24, 34) or at examining the behavior of wild-type or mixed

populations in chemostats (6, 9, 20, 46, 61, 65) showed thatcontinuous selection increases the proportion of mutators incell populations. In a previous paper, we demonstrated howeasily a population of Escherichia coli cells could become prin-cipally or totally mutators in response to several successiveselections (39). Here, we ask whether instead of changing me-dium in successive selections a single medium that selects di-rectly for mutator colonies could be devised. By asking cells toovercome several barriers to growth on a medium with limitingamounts of each required nutrient, we developed a plate me-dium on which only mutator cells grow. We have employed thismedium to examine the occurrence of mutators under severalconditions, and we discuss how this mimics the situation inmammalian cells that need to overcome several growth restric-tions before becoming proliferating cancer cells. We also showthat one mutator can induce a second mutator that then stim-ulates useful genetic changes, as part of a mutator cascade.

MATERIALS AND METHODS

Bacterial strains and plasmids. The strain CC107 carries an F9lacpro episomein the P90C (11, 43) strain background ara D(gpt-lac)5. The lac region on the Ffactor carries a lacI mutation and also a frameshift in the lacZ gene that revertsby the addition of a GC base pair to a monotonous run of GC base pairs. Thisstrain is described by Cupples et al. (11). AS18 is a Met2 derivative of CC107carrying an ICR-191-induced frameshift mutation in the metE (or possibly metR)gene. AS18-29 is a Bgl2 derivative of AS18 carrying an additional ICR-191-induced frameshift mutation, this time in the blgA gene. AS210 is a Leu2

derivative of AS18-29, carrying a mutH-induced frameshift mutation at the leulocus and a Tn10kan insert near the wild-type mutH gene (zgh-3159 [62]). Weconstructed specific mutator derivatives of certain strains by using P1 transduc-tion from strains in which a mini-Tn10 had integrated into either the mutT,mutH, mutS, mutL, or uvrD gene (45).

Genetic methods. Mapping experiments were carried out with P1 cotransduc-tion with Tn10 transposons that had integrated near either the mutH, mutL,mutS, or uvrD gene, or near various nutritional markers (62). Auxotrophs weredetected after ICR-191 mutagenesis by replicating Luria-Bertani (LB) platesspread with 100 to 300 colonies onto minimal medium plates and recognizingthose colonies which failed to grow. Combinations of supplements were used torestore growth, and then individual supplements were employed. All otherstrains and bacterial genetic methods, such as determination of rifampin resis-tance (Rifr), are described by Miller (43). We initially examined colonies forstrong mutator activity by gridding them onto an LB plate and growing themovernight at 32°C before replicating them onto a second LB plate. The second

* Corresponding author. Mailing address: Department of Microbi-ology and Molecular Genetics and The Molecular Biology Institute,University of California, Los Angeles, CA 90095. Phone: (310) 825-8460. Fax: (310) 206-3088. E-mail: [email protected].

1576

on February 13, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

plate was grown overnight, and then colonies were replicated onto an LB platewith 100 mg of rifampin per ml. After overnight growth, mutators defective inmismatch repair showed many Rifr colonies growing out of the replicated patch,whereas nonmutator strains did not. More quantitative measurements were thencarried out when relevant.

Preparation of cultures. Unless otherwise stated, all cultures for the experi-ments reported here were prepared by inoculating a portion of a single colonyinto LB medium (43) or other medium and growing it overnight. A differentsingle colony was used for each culture. Therefore, all mutants occurring indifferent cultures are of independent origin, since each single colony is derivedfrom a single isolated cell. Typically, 5-ml cultures were used.

Mutagenesis. All methods of mutagenesis were exactly as described in thework of Miller (43). 2-Aminopurine (2AP; Sigma) was used at concentrations of700 mg/ml. Cultures were prepared by subculturing 104 to 105 cells into 4 ml ofLB with 2AP, and these were grown for 12 to 16 generations in LB with 2APbefore plating. ICR-191 (Sigma) was used at concentrations of 10 mg/ml inminimal A medium (43) supplemented with 1 ml of LB broth, 2 ml of 20%glucose, 0.1 ml of 1 M MgSO4, and 0.5 ml of B1 (thiamine hydrochloride) per 100ml of medium.

Specialized media. Mutators were selected on lactose minimal A medium (43)supplemented with limiting amounts of required sugars or nutrients. We used150 mg of glucose per ml as a limiting carbon source and 0.5 mg of requiredamino acids, purines, or pyrimidines per ml. For instance, in conjunction withstrain AS18 or AS18-29 we used minimal A plates containing lactose, B1, andMgSO4 (33), supplemented with 150 mg of lactose per ml and 0.5 mg of methi-onine per ml. When called for, these plates were also supplemented with 40 mgof X-Glu (5-bromo-4-chloro-3-indolyl-b-D-glucoside; Research Organics) perml. This dye stains colonies with active b-glucosidase deep blue. We find thatcolonies of Bgl2 strains on plates containing glucose as a carbon source andX-Glu will yield blue papillae without the addition of another substrate forb-glucosidase. Thus, in the blue papillation method (47) as originally describedfor the lac system, phenyl-b-D-galactoside is added as a carbon source for papil-lae stained blue by X-Gal (5-bromo-4-chloro-3-indolyl-b-D-galactopyranoside).Here, this second carbon source is not necessary. Perhaps as a result of the lowerconcentration of the X-Glu in the role of carbon source, sugars are utilized in thefollowing order by revertants of the Lac2 Bgl2 strains: glucose . lactose .X-Glu.

Determination of mutator frequency. The data from Tables 3 and 4 were usedto calculate the spontaneous frequency of mismatch-repair-deficient mutants.Since the majority of cultures had no mutators, the fraction of cultures with nomutators was used to calculate the mean, m, for the Poisson distribution describ-ing the distribution of mutants. In the case of cells at a density of 105 cells perplate, the zero fraction is 327/355 5 0.921. Since for the Poisson distribution Po5 e2m, 0.921 is e2m, and m is 0.082, or, in this case, 0.082 3 1025. Accountingfor the 10% plating efficiency gives a value of m as 0.82 3 1025. Similar calcu-lations for cells at a density of 104 cells per plate give 0.63 3 1025, for a platingefficiency of 100%. We find that the plating efficiency at 104 cells varies somewhatmore than that for 105 cells, so we take 0.8 3 1025 as a more probable value. Thisvalue is for the fraction of mutators (or mutant frequency) in a culture grown forabout 33 generations.

Plating efficiency of mutators. Strain AS18-29 was plated at different densitieson selective medium, and then dilutions of a mutH derivative were plated on thesame plates and the percentages of mutator cells forming colonies were deter-mined.

RESULTS

Finding mutants with useful selective markers. We lookedfor mutations which could be used in combination with othermutations to provide useful selections for mutator strains. Weinitially targeted mutators lacking the mismatch repair system.These strains have greatly enhanced rates of transitions andalso of frameshifts at runs of single base pairs or repeatingdinucleotides (11, 33, 57). Therefore, strains such as CC107(11), which reverts from Lac2 to Lac1 via the addition of a Gto a run of six G’s, are useful for selecting for mutators, as wedescribed previously (39). In order to find additional markersto use in concert with Lac2 strains such as CC107, we mu-tagenized CC107 and derivatives of CC107 and looked forauxotrophs and then characterized the mutants with regard toreversion rates in both wild-type and mismatch-repair-deficient(mutH or mutS) strain backgrounds. We also determined thenutritional requirement and mapped the mutation to one ofthe known loci on the chromosome in most cases (see Mate-rials and Methods).

Auxotrophs found after treatment with ICR-191 are usuallycaused by additions or deletions at monotonous runs of GC

base pairs (5). Some of these should create monotonous runsof 6 bp or more and thus show greatly enhanced reversion ratesin a mismatch-repair-deficient strain compared to reversion ina wild-type strain. Approximately 1% of the survivors of ICR-191 treatment were auxotrophs, and about 75% of these couldbe identified with respect to the nutritional requirement. Mu-tants that were leaky or which did not show a significantlyincreased reversion in a mutH strain were discarded. Of 70mutants derived by ICR-191 mutagenesis, 22 merited furtherstudy. These mutants are shown in Table 1. Many of thesemutants have low reversion rates but show greatly increasedrates in a mutH background. Thus, they represent promisingindicators for mutators. Several of the strains, including AS18,which carries a mutation in the metE (or metR) gene, wereselected for additional experiments.

Detection of mutations in the bgl operon. As an additionalindicator of mutator strains, we sought mutations in the bgloperon that revert in response to mismatch-repair-defectivebackgrounds. Since the bgl operon is cryptic in most E. coliK-12 strains (56, 58), we first scored spontaneous mutants ofboth CC107 and AS18 that could grow on salicin as a solecarbon source for intense blue color on glucose minimal plateswith X-Glu (see Materials and Methods). We selected oneBgl1 mutant from each strain, mutagenized them with ICR-191, and screened for white colonies on X-Glu plates. Thesewere then tested for enhanced reversion in a mutH strain in avariation of the blue papillation assay (47), adapted for X-Glu(see Materials and Methods). One Bgl2 mutant was selectedfor further use. The mutant derived from CC107 was termedCC107-17, and the mutant derived from AS18 was namedAS18-29. In each case, the frequency of revertants to Bgl1 goesfrom approximately 10 per 108 cells in a wild-type strain to

TABLE 1. Mutants detected after ICR-191mutagenesis of strain CC107

Strain Requirement Locusa

No. of revertants/108 cellsb

WT mutH or mutS

AS1 Methionine metBFJL 30 ;700AS2 Adenosine 2 ;500AS3 ? 10 ;10,000AS4 Cysteine 2 ;500AS8 Adenine 3 ;2,000AS10 Leucine leu 46 ;2,000AS11 Guanine, adenosine 5 ;1,000AS18 Methionine metER 10 ;1,000AS101 Methionine metER 15 ;1,000AS107 Uracil, cytosine pyrE 30 ;10,000AS112 Adenosine purA 59 ;8,000AS114 Threonine thr 33 ;2,500AS115 Serine serA 6 ;1,500AS121 Guanine, adenosine purDH 3 ;5,000AS129 Histidine his 0 ;1,000AS130 Tryptophan trp 7 ;5,000AS131 Guanine, adenosine purDH 26 ;25,000AS132 Guanine, adenosine purF 2 ;10,000AS137 Methionine metBFJL 8 ;1,000AS139 Tryptophan trp 25 ;4,000AS141 Tyrosine tyr 7 ;3,500AS150 Methionine, cysteine 20 ;20,000

a Mutational locus responsible for the auxotrophic phenotype when deter-mined.

b The revertants per 108 cells were determined in multiple cultures. In the caseof mutH or mutS derivatives, the number of revertants is difficult to determineexactly, since colonies continue to arise after plating. Therefore, the numbersgiven should be considered estimates. WT, wild type.

VOL. 181, 1999 DIRECT SELECTION FOR MUTATORS IN ESCHERICHIA COLI 1577


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

approximately 5,000 per 108 cells in a mutH strain. Thesestrains form numerous blue papillae in a mismatch-repair-deficient background on X-Glu plates and thus serve as indi-cators for mutHLS and uvrD strains.

Sequential selection on a single medium detects mutators.We initially used strain AS18-29, which carries frameshift mu-tations in both lacZ and metE, to detect mutators. We exper-imented with different media to select colonies of strainAS18-29 that overcome both the Lac2 and the Met2 defects.We applied the principles depicted in Fig. 1 and 2. Lactosemedium with limiting amounts of glucose and methionineshould allow cells to grow to a small population size of approx-imately 105 cells before exhausting the limiting carbon source(glucose). To grow further would require metabolizing thelactose, and this would require a mutation from Lac2 to Lac1.

Then, the Lac1 cell could grow until it exhausted the methio-nine and again reach 105 cells. Now, a second mutation toMet1 would be required to allow further colony growth. Incases where a mutator population of 105 cells would have aLac1 or a Met1 revertant, we would expect each mutator cellto give rise to a full colony under the right conditions, whereasonly a fraction (about 1025 to 1024) of the nonmutator cellswould form colonies (Fig. 1). We therefore examined the plat-ing efficiency of a mutH derivative of AS18-29 on medium withdifferent limiting amounts of glucose and methionine andfound the best results with 150 mg of glucose per ml and 0.5 mgof methionine per ml (see Materials and Methods). On theseplates, the mutH strain formed colonies with a 50 to 100%plating efficiency, while the wild-type derivative formed nocolonies at all. We determined mutator colonies by eithertesting for enhanced frequency of Rifr colonies or includingX-Glu in the medium to indicate frequent mutation to Bgl1 bythe appearance of many blue papillae. (In the absence ofX-Glu, colonies are usually picked after 3 days, whereas in thepresence of X-Glu an extra day is needed for optimal papillaeformation.)

To determine whether the selective medium could detectmutators from a mixed population, we grew AS18-29 overnightin 2AP and first measured the mutators by direct visualizationon minimal glucose plates supplemented with methionine andX-Glu, as well as on the selective plates. The glucose–X-Gluplates with methionine allow the growth of all colonies, withmutators displaying a large number of blue papillae. This canbe seen in Fig. 3A, which shows that of several hundred colo-nies, one prominent mutator is evident. In these experiments,mutators deficient in mismatch repair are found in between1/1,000 and 1/300 cells after 2AP treatment. Figure 3B showsthe result of plating an even larger sample of cells on theselective medium. Now, only the mutator colonies grow, asindicated by the blue color of the colonies, which is, uponmagnification (Fig. 3C), really due to thousands of Bgl1 pa-pillae from the X-Glu indicator. Figure 3C also allows thevisualization of the microcolonies that failed to develop intotrue colonies. We verified that the surviving colonies werepredominantly mutators by testing for increased frequency ofRifr mutants. Table 2 shows the results from several experi-ments. It can be seen that approximately 90% of the full-sizecolonies that grow on the selective plates after 2AP mutagen-esis of strains AS18 and AS18-29 are strong mutators. Becauseboth mutations that need to revert to restore the Lac1 Met1

phenotype are frameshifts, it is expected that the mutators aredeficient in the mismatch repair system. We have examinedsample mutators and found that all of the mutants tested carrya mutation in one of the four mismatch repair loci (mutHLS oruvrD [see below]).

We have carried out reconstruction experiments by mixingdifferent proportions of wild-type and mutH derivatives ofAS18-29 and then plating them on selective medium. As Table3 shows, the selective plates are sensitive to cell density. Thecell lawn eats up some of the limiting nutrients, and the effi-ciency of recovering a mutator goes down as the cell densityincreases. Placing up to 10,000 cells on a plate gives optimalresults. As the number of cells approaches and then exceeds105 per plate, the efficiency of recovery of mutators dropssignificantly. With 105 cells per plate, the efficiency drops from50 to 100% to 10%, and with 106 cells, the efficiency drops to1% recovery. Despite this, we can still detect mutators thatoccur spontaneously, at levels near 1 per 105 cells (see below).

Detection of spontaneous mutators. We employed strainAS18-29 to detect spontaneous mutators in cultures grown forthree generations. We examined 355 cultures, by plating 105

FIG. 1. Selection for mutator colonies. Lac2 Met2 cells are plated on lactosemedium with limiting amounts of glucose and methionine. Cells form a micro-colony before exhausting the glucose. Revertants to Lac1 can grow further. Only1% of the nonmutator microcolonies (left) will have a Lac1 cell, whereas 100%of the mutator microcolonies (right) will have a Lac1 cell. Further growth of theLac1 cells yields microcolonies that exhaust the methionine. Again, 1% of thenonmutator Lac1 microcolonies (left) will have a Met1 cell, whereas 100% ofthe mutator Lac1 microcolonies (right) will. The very rare nonmutator Lac1

Met1 colonies have no Bgl1 papillae (left), but the Lac1 Met1 mutator coloniesshow microsatellite instability and give Bgl1 papillae (right; black dots). WT,wild type.

FIG. 2. A representation of mutator colonies growing to full size after theselection described in the legend to Fig. 1. Many small microcolonies (open dots)are seen in the background, but only two full-grown colonies appear, and thesehave many Bgl1 papillae (dark dots).

1578 MILLER ET AL. J. BACTERIOL.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

cells from each culture, and an additional 344 cultures, byplating 104 cells from each culture. Colonies were detected onselection medium without X-Glu and purified once beforebeing tested for the frequency of Rifr colonies. Because thefrequency of spontaneous mutators is on the order of 1025 inthe population, some Lac1 Met1 colonies are due to sponta-neous mutations occurring in nonmutator cells. These coloniesalso appear on the order of 1025 in the population. As Table4 shows, about 50% of the colonies detected were mutators.About 92% of the plates with 105 cells yielded no mutatorcolonies. Applying the 10% plating efficiency determined fromreconstruction experiments (Table 3), we can calculate themutator frequency from the fraction of cultures with no mu-tators. This gives a frequency of mutators of 0.8 3 1025 (seeMaterials and Methods). For the cultures plated with only 105

cells per plate, which gives a plating efficiency of 50 to 100%,approximately 94% of the cultures yielded no mutator colo-nies, which translates to 0.6 3 1025 to 1.2 3 1025 for themutator frequency. These values are close to estimates basedon other measurements we have made, as described in theDiscussion.

Identification of mutators detected spontaneously. Wemapped the mutation causing the mutator phenotype in 45 of

the spontaneous mutators detected in the above experiments,by P1 transduction (see Materials and Methods). All of themutations fell into one of the four mismatch repair genes.However, as Table 5 shows, the mutations were not distributedequally among mutL, mutH, mutS, and uvrD. Instead, 19 of themutations were in mutH, 18 were in mutL, 7 were in mutS, and1 was in uvrD.

Different combinations of markers are effective in selection.The strain AS18-29 carries a mutation in the lac region on anF9 plasmid and a met mutation in the chromosome. The selec-tive medium is not limited to those markers, nor does it dependon having a mutation on an F9 plasmid (data not shown).Several different combinations of markers were made for usein the successive selection medium. Some of the markers arederived from the experiment whose results are shown in Table1. A second set of markers was derived by first making a mutHderivative of AS18-29 and obtaining auxotrophs (in the pres-ence of methionine) induced by the mutator effect of the mutHallele. Then, the mutH allele was crossed out, and the reversionrates were compared. The new strains from this selection carrymutations that are not limited to additions or deletions atmonotonous runs of G’s or C’s. They may contain additions ordeletions at runs of A’s or T’s, or at repeating dinucleotides, aswell as certain base substitutions. Table 6 shows the useful

TABLE 3. Efficiency of plating of mutH derivative of AS18-29 inthe presence of different densities of cells of AS18-29a

Strain Value/plate

AS18-29 (no. of cells) 106 105 104

mutH derivative (% plating efficiency) 1 10 50–100

a See Materials and Methods for details.

FIG. 3. Selection for mutators. (A) Strain AS18-29 was grown overnight in LB broth with 2AP and plated on X-Glu plates. Here, one mutator colony (see insert)is evident, as indicated by the numerous blue Bgl1 papillae. (B) A photograph of a plate like that diagrammed in Fig. 2. Two full-size mutator colonies grow on selectiveplates onto which several thousand cells have been spread. Minute microcolonies in the background are nonmutator colonies. The mutator colonies appear blue becauseof thousands of blue Bgl1 papillae (see magnification in panel C). (C) A magnification of part of the plate shown in panel B. Here, the nonmutator microcolonies areevident, as are the blue Bgl1 papillae, representing the microsatellite instability of the full-size mutator colony.

TABLE 2. Percentage of mutators among Lac1 Met1 colonies in2AP cultures of AS18 and AS18-29 after the plate selection

described in the text and for Fig. 1 to 5

Strain No. of 2APcultures tested

No. of coloniestested

No. of mutatorcolonies

%Mutator

AS18 5 91 81 89AS18-29 5 136 127 93



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

strains obtained from this selection. One additional mutationused in some experiments (data not shown) is the argE ambermutation derived from strain XAC (43). This mutation revertsvia base substitutions that either restore the UAG codon to asense codon or else create an amber or ocher suppressor.

Triple selection. We utilized strain AS210 (Table 6), whichcarries frameshift mutations in lacZ, metE, and leu. Eventhough three successive selections occur, colonies were ob-tained on plates with limiting amounts of glucose, methionine,and leucine. After mutagenesis with 2AP, 97% of the colonies(26 of 27) were found to be mutators, and 90% of the coloniesdetected spontaneously were mutators. All of the mutator col-onies tested had defects in one of the mismatch repair genes.The plating efficiency of a mutator on this medium, however,does not exceed 10%.

Mutator cascade. Some mutators may not exhibit an in-creased rate of mutations at certain sequences, such as frame-shifts at runs of identical bases of repeated sequences. How-ever, they might show an increased rate of mutation at asecond mutator locus, the resulting secondary mutators nowbeing able to stimulate the specific sequence change. This typeof mutator cascade might be significant in creating certainphenotypes. We can examine this phenomenon by looking atour tester strain, AS18-29, into which we have crossed a mu-tation at the mutT locus. The resulting mutators show an in-creased incidence of only one specific transversion, A:T3C:G(10), because of the failure to hydrolyze the oxidatively dam-aged DNA synthesis precursor 8-oxo-dGTP (38). That mutTcells do not greatly increase mutations at the frameshifts weused to monitor mutations in AS18-29 is evidenced in Table 7.Comparing the wild-type and mutT derivatives of AS18-29 withrespect to reversion of either the lac, met, or bgl frameshiftmutation shows either only slight or no differences comparedwith the mutH derivative of AS18-29, which shows an enor-mous increase (Table 7). Yet, when we look at the incidence ofmutations in the mutHLS pathway that create mutators thatcan now stimulate the frameshifts in lac, met, and bgl, we see

that mutT increases the spontaneous level of these mutatorsfrom 8 3 1026 to 3 3 1024.

In the mutT derivative, approximately 1 of each 3,300 cellshas a mutation in the mutHLS system. As a single cell dividesand forms a colony, on average at some point between the 12thand 16th cell division the first mutator arises and forms a verythin sector, representing the lineage of daughter cells in thecolony. Occasionally, a mutHLS mutator arises in the first celldivision, yielding a colony that is half mutT alone and halfmutT mutH (or mutL or mutS). We can visualize these sectorsby decorating them with blue papillae arising from the Bgl1

reversion event at the frameshift in the bglA gene in AS18-24.Figure 4 shows some of these sectors. Recall that the startingmutT derivative of AS18-29 cannot give rise to frequent Bgl1

papillae. Figure 4A shows a colony that is half sectored for themutHLS phenotype. Figure 4B shows a colony with a smallermutator sector.

When the subpopulation of mutT-generated mutHLS muta-tors within a colony sector reaches a sufficient size, it can giverise to the first revertant able to overcome one of the growthrestrictions imposed by the Lac2 Met2 Bgl2 phenotype of theAS18-29 strain background. As depicted in Fig. 1 and 2, thiscan then lead to a progression that leads to the development ofa full-size mutator colony. As Fig. 5 diagrams, in the caseshown here, the starting strain can grow to only a certain sizeon medium with limiting methionine and limiting glucose.Then, the mutHLS mutator cell sector generates a Lac1 re-vertant (for instance), which can grow further, and then aMet1 cell that can grow without restrictions arises. The rapidlygrowing colony thus arises from within a slow-growing colony,extends out beyond the original colony, and displays microsat-ellite instability by throwing off frequent blue papillae in thepresence of X-Glu. Figure 6 shows some of these colonies.Figure 6A and B show single colonies outgrowing the originalslow-growing colony, and Fig. 6C shows how the unrestricted

TABLE 4. Spontaneous mutators in cultures of AS18-29 detectedon selection plates with different cell densitiesa

No. ofAS18-29

cells

No. ofcultures

No. of cul-tures with no

mutators

Mutatorplating

efficiency(%)

No. ofmutators

found

No. ofnonmu-tators

% Muta-tors

105 355 327 10 43 48 47104 344 323 50–100 21 27 44

a Close to 700 cultures were plated at two different cell densities, and thenumber of mutators found in samples of each culture was recorded. Each samplerepresented either 104 or 105 cells. The frequency of mutators in the populationwas determined from the fraction of cultures with no mutators and the platingefficiency at the respective cell density (see Materials and Methods).

TABLE 5. Position of the mutation resulting in the mutatorphenotype in each of the 45 independent mutators

of spontaneous origin

Mutated gene

No. of spontaneousmutators fromAS18-29 with

mutation

mutH ..................................................................................... 19mutL...................................................................................... 18mutS ...................................................................................... 7uvrD....................................................................................... 1

TABLE 6. Auxotrophic requirement of ICR-191-inducedmutants of strain AS18-29a

Strain Requirement Locus

No. of revertants/108 cells

WT mutH

AS201 Cysteine cys 300 ;10,000AS210 Leucine leu 40 ;5,000AS212 Isoleucine ile 700 ;10,000AS213 Arginine arg 20 ;10,000AS220 Adenine pur 5 ;10,000

a Each of the auxotrophic requirements is in addition to the Met2 phenotypeof AS18-29 (see legend to Fig. 1). WT, wild type.

TABLE 7. Mutant frequencies for wild type and derivativesa

StrainNo. of cultures with characteristic:

Lac1 Met1 Bgl1 MMR2

WT 79 6 4 8.3 6 3.8 6.9 6 4.3 ;800mutT 76 6 18 9.1 6 3.4 8.1 6 3.7 ;30,000mutH ;10,000 ;1,000 ;5,000

a Mutant frequencies determined as the averages of multiple cultures areshown for wild type (WT) (strain AS18-29; Lac2 Met2 Bgl2) and mutT or mutHderivatives prepared by transducing into AS18-29 a mini-Tn10 inserted intoeither the mutT or mutH locus. Mismatch repair deficiency (MMR2) resultsfrom loss of mutHLS or uvrD.



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

growth can ultimately invade other surrounding slow-growingor nongrowing microcolonies.

DISCUSSION

Mutator cells, defined as cells with higher rates of mutationthan those of wild type, were first detected in Drosophila mela-nogaster in the early 1940’s (54) and in bacteria in the early1950’s (68). Now, mutations in more than 20 different genes inbacteria such as E. coli have been shown to result in themutator phenotype (see reviews in references 8, 13, 19, 44, and64). Most of these cause defects in repair or damage avoidancesystems, although some affect less well defined pathways (7, 42,63). Higher cells also display mutator phenotypes when certaingenes are inactivated, and this can lead to increased incidenceof disease. For instance, inheriting one copy of a defective geneinvolved in the human counterpart to the bacterial mismatchrepair system leads to increased susceptibility to colon cancer,HNPCC-related endometrial cancer, and ovarian cancer (4,18, 30, 49; see review in reference 37). The resulting tumorlines are mutators that show increased repeat-tract instability(see reviews in references 27 and 35) and higher mutation ratesin genes such as hprt (1, 3, 21). Loeb had already postulatedthat creation of a mutator cell would be an early step in somecases of carcinogenesis, since a mutator can generate the mul-tistep mutations faster than can a normal cell with lower mu-

tation rates (27, 36). How then do mutators arise in a popula-tion of cells, and how do they proliferate?

Clearly, spontaneous mutants constantly arise in a popula-tion, and some of these are mutators. The balance of selectiveadvantage and disadvantage will affect the proportion of mu-tators in any given environment, but the continued generationof new spontaneous mutations will provide a constant source.Using the selection system described in this work and discussedfurther below allows us to measure the proportion of cells in agrowing population of E. coli cells that have defects in one ofthe four mismatch repair genes. This number of just under1025 (0.8 3 1025) mutants is in a population growing in brothfor about 33 generations. This number is in agreement withestimates from our previous study (39), in which Lac1 frame-shift mutants detected at 2 3 1027 were scored to reveal that0.5% (5 3 1023) were mutators. These values predict thatmutants that are both Lac1 and defective in one of the fourmismatch repair loci are present at 1029 in the population (theproduct of these latter two frequencies). However, since theLac1 revertant rate in each mutator subpopulation is close to1024 (9, 30), then the mutators are estimated to be in thepopulation at 1029 divided by 1024, or close to 1025. (Inter-estingly, a subsequent analysis of mismatch repair mutators inSalmonella typhimurium shows that they occur spontaneouslyan order of magnitude less frequently than found here forE. coli [32]).

FIG. 4. Appearance of mutator subpopulations as colony sectors. A mutT derivative of strain AS18-29 was plated on minimal glucose plates with methionine andX-Glu. Mismatch-repair-deficient (MMR2) sectors (e.g., mutH) are revealed by frequent blue Bgl1 papillae (microsatellite instability). (A) An MMR2 mutant hasarisen at the first cell division, resulting in half of the colony having microsatellite instability. (B) An MMR2 mutant has arisen at approximately the fifth cell division,resulting in a thinner sector of the colony displaying microsatellite instability.



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

Looking at the distribution of mutations among the fourmismatch repair loci (Table 5) that leads to a combined rate of0.8 3 1025 mutators in the population makes us realize howthese measurements can vary, depending on the specific genebeing monitored. The mutant frequencies for each of the fourloci are as follows: mutH, 3.4 3 1026; mutL, 3.2 3 1026; mutS,1.2 3 1026; uvrD, 1.8 3 1027. It is not clear why only 1 of 45mutations is in the uvrD gene. Reconstruction experimentsindicate that the efficiency of plating of strains with uvrD mu-

tations in the selection employed is the same as for strains withmutations in mutH (data not shown), suggesting that the effectis due to rates of mutations themselves, but it is possible thatother hidden experimental biases conspire to reduce the ap-pearance of UvrD mutants.

We should note at this point that mutagenesis from any of anumber of sources serves to increase the proportion of muta-tors to near or above 1023 (see, for instance, reference 39).However, in addition to mutagenesis, selection for mutantphenotypes can serve to increase the frequency of mutators ina population. Chemostat experiments show that extensive con-tinued growth in a full-nutrient environment can still lead tothe selection of fitter strains and increase the proportion ofmutators, since they can give rise to fitter variants more rapidlythan can the wild type (6, 9, 20, 46, 61, 65). Computer simu-lations also argue that mutators are selected for in continuedgrowth in chemostats (67). In fact, some mutators were origi-nally found by procedures designed to screen mutagenizedcells after single or successive selections (12, 23, 24, 34) or byobservation of cells selected for a specific phenotype (60, 61).Several investigators have found that natural isolates of bac-teria have several percent mutators (22, 28, 31), suggesting thatpopulations in the wild might be undergoing constant selec-tion. Cebula and coworkers have argued that the several per-cent mutators found among E. coli and Salmonella strainsisolated from patients might be related to pathogenicity (31),although others have disputed this correlation (41).

In a previous study (39), we showed how quickly the pro-portion of mutators can rise in a population, and how severalsuccessive selections can result in the entire surviving popula-tion being mutator. This underscores the consequences of cer-tain chemotherapies, in which the resistant cells may be en-riched for mutators. In the previous work (39), the successiveselections involved transferring cells to different media. Wehave extended that study in the work reported here by design-ing a single medium that selects for several phenotypes in

FIG. 5. The effects of a mutator cascade are depicted in a schematic diagramof a mutT derivative of strain AS18-29 (Lac2 Met2 Bgl2) growing on lactoseminimal medium with trace amounts of glucose and methionine and the indica-tor X-Glu. The starting cell (A) forms a microcolony, in which a mutH cell arises(B). As the microcolony grows, exhausting the glucose in the medium, thesubpopulation of mutH cells also slowly proliferates (C), until a Lac1 cell ariseswithin the mutH subpopulation (D). The Lac1 cells can grow further on thelactose in the medium until the trace methionine is exhausted. If a Met1 cellappears (E), it can now grow without the restriction of limiting methionine orglucose, expanding rapidly and showing microsatellite instability by the appear-ance of many blue Bgl1 papillae (F).

FIG. 6. Mutator cascade. The figure shows selection for the growth of a mutator arising from a microcolony. The starting microcolony is a mutT derivative of strainAS18-29. The mutT mutator cannot revert the frameshift mutations in lacZ and metE, but can generate mutations in the mutH, -L, or -S genes, resulting inmismatch-repair-deficient (MMR2) mutators that can revert the frameshifts and break free of growth restrictions. (A and B) MMR2 mutator colonies growing outfrom a microcolony. The MMR2 mutators have many blue papillae, demonstrating microsatellite instability (see legend to Fig. 4). (C) An MMR2 mutator generatedon a more crowded plate will grow in an unrestricted fashion, proliferating over the plate and overrunning the nonmutator microcolonies.



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

succession. Cells with multiple nutrient requirements are chal-lenged to form colonies by eventually overcoming each of thegrowth barriers. In the most studied case, strain AS18-29, car-rying frameshifts in the lacZ gene on the F plasmid and in thechromosomal metE gene, must revert both mutations in orderto form full-size Lac1 Met1 colonies. A second strain adds athird growth requirement via a frameshift mutation in thechromosomal leu operon. The medium contains very smallamounts of glucose as a carbon source and small amounts ofmethionine (and also of leucine when relevant). This permits amicrocolony of nearly 105 cells to form, which if derived froma mutator cell will contain enough cells to have a mutant thatcan now use, for example, the lactose in the medium as acarbon source to initiate a new microcolony of 105 cells beforeit exhausts the methionine. A mutator cell will now have asecond mutation in this new microcolony that reverses themetE mutation, allowing it to form a full-size colony (in thecase of the derivative with the mutation in leu, a third round ofsuccessive microcolony formation would be required). Full-sizecolonies can be picked and analyzed or directly visualized byusing reversion of a frameshift mutation in the bgl operon todecorate mutators with scores of blue papillae. The results,shown in Fig. 3, reveal that cells having to overcome severalgrowth requirements in succession on a single defined mediumgive rise to mutator colonies. After mild mutagenesis, 90 to100% of the colonies on this medium are mutators with defectsin the mismatch repair system, depending on whether two orthree growth requirements are employed. Even without a mu-tagen, spontaneous mutators constitute 50% of the colonies ofcells that break through the growth restrictions.

It is interesting to compare the emergence of mutator col-onies on plates selecting for overcoming several successivegrowth restrictions with the emergence of a cancer cell thatundergoes successive mutations to break free of growth restric-tions, particularly in the case of colon cancer. In both cases,frameshift mutations at runs of mono- or dinucleotides areinvolved in creating mutants. In the case of the bacterial strainshown here, the frameshift mutations restore the normal gene,whereas in the carcinogenesis model, frameshifts inactivatecertain genes. Many colon cancer cell lines have mutations inthe APC gene (25), the rII gene encoding the negative growthsuppressor transforming growth factor b II (40, 50), and theapoptosis-associated BAX gene (55). The APC gene containsruns of A residues and an AG dinucleotide repeat that areframeshifted in sequenced mismatch-repair-deficient tumorlines (25). In these lines, most of the mutations inactivatingtransforming growth factor b are at a run of 10 A’s or at athreefold repeat of a GT sequence in the rII gene (40), andframeshifts at a run of eight G’s are found in the BAX gene(55). It is easy to see the parallels between successive selectionsthat enrich for mismatch-repair-deficient strains in bacteriaand those in human tumor lines that have to overcome severalgrowth restrictions.

Figures 4 to 6 portray the events that occur when a mutHLSmutator cell arises within a colony of cells growing very slowlyunder restrictive conditions. As all the cells grow, the patch ofcells derived from the mutHLS mutator cell (Fig. 4) reaches asufficient size to allow the appearance of a mutant that canovercome subsequent growth restrictions (Fig. 5), proliferateduring unrestricted growth, and exhibit repeat-tract instability(Fig. 6). In this series of experiments, the appearance of themutHLS mutator is accelerated by the presence of a differentmutator, in this case mutT, that cannot revert the frameshifts(Table 7) to overcome the growth restrictions but that cangenerate mutations that inactivate the mutH, mutL, or mutSgene. This produces a mutator cascade, where one mutator

induces a second one. We have also found similar results withthe mutA mutators, which result from miscoding tRNAs (datanot shown). It will be interesting to see whether any examplesof such a mutator cascade are found among tumor lines, orwhether any cancer susceptibilities are found to result from theinheritance of a defective copy of a different repair gene thatby itself leads to a mutator effect without microsatellite insta-bility but which can generate mismatch-repair-deficient muta-tors. There is precedent for inactivation of the mismatch repairgenes as a second step in the development of colon cancer. Asignificant fraction of sporadic colon cancer tumor lines thatshow microsatellite instability suffer inactivation of mismatchrepair genes as a consequence of the epigenetic gene silencingproduced by hypermethylation of both copies of the relevantpromoters (29).

ACKNOWLEDGMENT

This work was supported by grant GM32184 to J.H.M. from theNational Institutes of Health.

REFERENCES

1. Aaltonen, L. A., P. Peltomaki, F. S. Leach, P. Sistonen, L. Pylkkanen, J.-P.Mecklin, H. Jarvinen, S. M. Powell, J. Jen, S. R. Hamilton, G. M. Petersen,K. W. Kinzler, B. Vogelstein, and A. de la Chapelle. 1993. Clues to thepathogenesis of familial colorectal cancer. Science 260:812–816.

2. Armitage, P., and R. Doll. 1954. The age distribution of cancer and multi-stage theory of carcinogenesis. Br. J. Cancer 8:1–12.

3. Bhattacharyya, N. P., A. Skandalis, A. Ganesh, J. Groden, and M. Meuth.1994. Mutator phenotypes in human colorectal carcinoma cell lines. Proc.Natl. Acad. Sci. USA 91:6319–6323.

4. Bronner, C. E., S. M. Baker, P. T. Morrison, G. Warren, L. G. Smith, M. K.Lescoe, M. Kane, C. Earabino, J. Lipford, A. Lindblom, P. Tannergard,Bollag, A. R. Godwin, D. C. Ward, M. Nordenskjold, R. Fishel, R. Kolodner,and R. M. Liskay. 1994. Mutation in the DNA mismatch repair gene homo-logue hMLH1 is associated with hereditary nonpolyposis colon cancer. Na-ture 368:258–261.

5. Calos, M. P., and J. H. Miller. 1981. Genetic and sequence analysis offrameshift mutations induced by ICR-191. J. Mol. Biol. 153:39–66.

6. Chao, L., and E. C. Cox. 1983. Competition between high and low mutatingstrains of Escherichia coli. Evolution 37:125–134.

7. Connolly, D. M., and M. E. Winkler. 1989. Genetic and physiological rela-tionships among the miaA gene, 2-methylthio-N6-(D2-isopentenyl)-adeno-sine tRNA modification, and spontaneous mutagenesis in Escherichia coliK-12. J. Bacteriol. 171:3233–3246.

8. Cox, E. C. 1976. Bacterial mutator genes and the control of spontaneousmutation. Annu. Rev. Genet. 10:135–156.

9. Cox, E. C., and T. C. Gibson. 1974. Selection for high mutation rates inchemostats. Genetics 77:169–184.

10. Cox, E. C., and C. Yanofsky. 1967. Altered base ratios in the DNA of anEscherichia coli mutator strain. Proc. Natl. Acad. Sci. USA 58:1895–1902.

11. Cupples, C. G., M. Cabrera, C. Cruz, and J. H. Miller. 1990. A set of lacZmutations in Escherichia coli that allow rapid detection of specific frameshiftmutations. Genetics 125:275–280.

12. Degnen, G. E., and E. C. Cox. 1974. A conditional mutator gene in Esche-richia coli: isolation, mapping, and effector studies. J. Bacteriol. 117:477–487.

13. Drake, J. W. 1991. A constant rate of spontaneous mutation in DNA-basedmicrobes. Proc. Natl. Acad. Sci. USA 88:7160–7164.

14. Drake, J. W. 1991. Spontaneous mutation. Annu. Rev. Genet. 25:125–146.15. Drake, J. W. 1993. Rates of spontaneous mutation among RNA viruses.

Proc. Natl. Acad. Sci. USA 90:4171–4175.16. Eigen, M. 1993. Viral quasispecies. Sci. Am. 269:42–49.17. Eshleman, J. R., and S. D. Markowitz. 1995. Microsatellite instability in

inherited and sporadic neoplasms. Curr. Opin. Oncol. 7:83–89.18. Fishel, R., M. K. Lescoe, M. R. S. Rao, N. G. Copeland, N. A. Jenkins, J.

Garber, M. Kane, and R. Kolodner. 1993. The human mutator gene homologMSH2 and its association with hereditary nonpolyposis colon cancer. Cell 75:1027–1038.

19. Friedberg, E. C., G. C. Walker, and W. Seide. 1995. DNA repair and mu-tagenesis. American Society for Microbiology, Washington, D.C.

20. Gibson, T. C., M. L. Scheppe, and E. C. Cox. 1970. Fitness of an Escherichiacoli mutator gene. Science 169:686–688.

21. Glaab, W. E., and K. R. Tindall. 1997. Mutation rate at the hprt locus inhuman cancer cell lines with specific mismatch repair-gene defects. Carci-nogenesis 18:1–8.

22. Gross, M. D., and E. C. Siegel. 1981. Incidence of mutator strains in Esch-erichia coli and coliforms in nature. Mutat. Res. 91:107–110.

23. Helling, R. B. 1968. Selection of a mutant of Escherichia coli which has high



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

mutation rates. J. Bacteriol. 96:975–980.24. Hoess, R. H., and R. K. Herman. 1975. Isolation and characterization of

muator strains of Escherichia coli K-12. J. Bacteriol. 122:474–484.25. Huang, J., N. Papadopoulos, A. J. McKinley, S. M. Farrington, L. J. Curtis,

A. H. Wyllie, S. Zheng, J. K. V. Willson, S. D. Markowitz, P. Morin, K. W.Kinzler, B. Vogelstein, and M. G. Dunlop. 1996. APC mutations in colorectaltumors with mismatch repair deficiency. Proc. Natl. Acad. Sci. USA 93:9049–9054.

26. Ionov, Y., M. A. Peinado, S. Malkhosyan, D. Shibata, and M. Perucho. 1995.Ubiquitous somatic mutations in simple repeated sequences reveal a newmechanism for colonic carcinogenesis. Nature 363:558–561.

27. Jackson, A. L., and L. A. Loeb. 1998. The mutation rate and cancer. Genetics148:1483–1490.

28. Jyssum, K. 1960. Observations on two types of genetic instability in Esche-richia coli. Acta Pathol. Microbiol. Scand. 48:113–120.

29. Kane, M. F., M. Loda, G. M. Gaida, J. Lipman, R. Mishra, H. Goldman,J. M. Jessup, and R. Kolodner. 1997. Methylation of the hMLH1 promotercorrelates with lack of expression of hMLH1 in sporadic colon tumors andmismatch repair-defective human tumor cell lines. Cancer Res. 57:808–811.

30. Leach, F. S., N. C. Nicolaides, N. Papadopoulos, B. Liu, J. Jen, R. Parsons,P. Petlomaki, P. Sistonene, L. A. Aaltonen, M. Nystrom-Lahti, X.-Y. Guan,J. Zhang, P. S. Meltzer, J.-W. Yu, F.-T. Kao, D. J. Chen, K. M. Cerosaletti,R. E. Fournier, S. Rodd, T. Lewis, R. J. Leach, S. L. Naylor, J. Weissenbach,J.-P. Mecklin, H. Jarvinen, G. M. Petersen, S. R. Hamilton, J. Green, J. Jass,P. Watson, H. T. Lynch, J. M. Trent, A. de la Chapelle, K. W. Kinzler, andB. Vogelstein. 1993. Mutations of a mutS homolog in hereditary nonpolypo-sis colorectal cancer. Cell 75:1215–1225.

31. LeClerc, J. E., B. Li, W. L. Payne, and T. A. Cebula. 1996. High mutationfrequencies among Escherichia coli and Salmonella pathogens. Science 274:1208–1211.

32. LeClerc, J. E., W. L. Payne, E. Kupchella, and T. A. Cebula. 1998. Detectionof mutator subpopulations in Salmonella typhimurium LT2 by reversion ofhis alleles. Mutat. Res. 400:89–97.

33. Leong, P. M., H. C. Hsia, and J. H. Miller. 1986. Analysis of spontaneousbase substitutions generated in mismatch-repair-deficient strains of Esche-richia coli. J. Bacteriol. 168:412–416.

34. Liberfarb, R. M., and V. Bryson. 1970. Isolation, characterization and ge-netic analysis of mutator genes in Escherichia coli B and K-12. J. Bacteriol.104:363–375.

35. Loeb, L. A. 1991. Mutator phenotype may be required for multistage carci-nogenesis. Cancer Res. 51:3075–3079.

36. Loeb, L. A. 1994. Microsatellite instability: marker of mutator phenotype incancer. Cancer Res. 54:5059–5063.

37. Lynch, H. T., T. Smyrk, and J. Lynch. 1997. An update of HNPCC (Lynchsyndrome). Cancer Genet. Cytogenet. 93:84–99.

38. Maki, H., and M. Sekiguchi. 1992. MutT protein specifically hydrolyses apotent mutagenic substrate for DNA synthesis. Nature 355:273–275.

39. Mao, E. F., L. Lane, J. Lee, and J. H. Miller. 1997. Proliferation of mutatorsin a cell population. J. Bacteriol. 179:417–422.

40. Markowitz, S., J. Wang, L. Myeroff, R. Parsons, L. Sun, J. Lutterbaugh, R. S.Fan, E. Zborowska, K. W. Kinzler, B. Vogelstein, M. Brattain, and J. K. V.Willson. 1995. Inactivation of the type II TGF-b receptor in colon cancercells with microsatellite instability. Science 268:1336–1338.

41. Matic, I., M. Radman, F. Taddei, B. Picard, C. Doit, E. Bingen, E. Denamur,and J. Elion. 1997. Highly variable mutation rates in commensal and patho-genic Escherichia coli. Science 277:1833–1834.

42. Michaels, M. L., C. Cruz, and J. H. Miller. 1990. mutA and mutC; twomutator loci in Escherichia coli that stimulate transversions. Proc. Natl.Acad. Sci. USA 89:9211–9215.

43. Miller, J. H. 1992. A short course in bacterial genetics: a laboratory manualand handbook for Escherichia coli and related bacteria, p. 194–195. ColdSpring Harbor Laboratory Press, Cold Spring Harbor, N.Y.

44. Miller, J. H. 1996. Spontaneous mutators in bacteria: insights into pathwaysof mutagenesis and repair. Annu. Rev. Microbiol. 50:625–643.

45. Miller, J. H. Unpublished data.46. Nestman, E. R., and R. F. Hill. 1973. Population changes in continuously

growing mutator cultures of Escherichia coli. Genetics 73(Suppl.):41–44.47. Nghiem, Y., M. Cabrera, C. G. Cupples, and J. H. Miller. 1988. The mutY

gene: a mutator locus in Escherichia coli that generates GC3TA transver-sions. Proc. Natl. Acad. Sci. USA 85:2709–2713.

48. Nowell, P. C. 1976. The clonal evolution of tumor cell populations. Science194:23–28.

49. Papadopoulos, N., N. C. Nicolaides, Y. F. Wei, S. M. Ruben, K. C. Carter,C. A. Rosen, W. A. Haseltine, R. D. Fleischmann, C. M. Fraser, and M. S.Adams. 1994. Mutation of a mutL homolog in hereditary nonpolyposis coloncancer. Science 263:1625–1629.

50. Parsons, R., G.-M. Li, M. J. Longley, W.-H. Fang, N. Papadopoulos, J. Jen,A. de la Chapelle, K. W. Kinzler, B. Vogelstein, and P. Modrich. 1993.Hypermutability and mismatch repair deficiency in RER tumor cells. Cell 75:1227–1236.

51. Parsons, R., L. Myeroff, B. Liu, J. K. V. Willson, S. D. Markowitz, K. W.Kinzler, and B. Vogelstein. 1995. Microsatellite instability and mutations ofthe transforming growth factor beta type II receptor gene in colorectalcancer. Cancer Res. 55:5548–5550.

52. Peinado, M. A., S. Malkhosyan, A. Velazquez, and M. Perucho. 1992. Isola-tion and characterization of allelic losses and gains in colorectal tumors byarbitrarily primed polymerase chain reaction. Proc. Natl. Acad. Sci. USA 89:10065–10069.

53. Peltomaki, P., R. A. Lothe, L. A. Aaltonen, L. Pylkkanen, M. Nystrom-Lahti,R. Seruca, L. David, R. Holm, D. Ryberg, A. Gaugen, A. Brogger, A.-L.Borresen, and A. de la Chapelle. 1993. Microsatellite instability is associatedwith tumors that characterize the hereditary non-polyposis colorectal carci-noma syndrome. Cancer Res. 53:5853–5855.

54. Plough, H. H. 1941. Spontaneous mutability in Drosophila. Cold SpringHarbor Symp. Quant. Biol. 9:127–137.

55. Rampino, N., H. Yamamoto, Y. Ionov, Y. Li, H. Sawai, J. C. Reed, and M.Perucho. 1997. Somatic frameshift mutations in the BAX gene in coloncancers of the microsatellite mutator phenotype. Science 275:967–969.

56. Reynolds, A. E., J. Felton, and A. Wright. 1981. Insertion of DNA activatesthe cryptic bgl operon in E. coli K12. Nature 293:625–629.

57. Schaaper, R. M., and R. L. Dunn. 1987. Spectra of spontaneous mutations inEscherichia coli strains defective in mismatch correction: the nature of in vivoDNA replication errors. Proc. Natl. Acad. Sci. USA 84:6220–6224.

58. Schaefler, S. 1967. Inducible system for the utilization of betaglucosides inEscherichia coli. I. Active transport and utilization of beta-glucosides. J.Bacteriol. 93:254–263.

59. Shibata, D., W. Navidi, R. Salovaara, Z. H. Li, and L. A. Aaltonen. 1996.Somatic microsatellite mutations as molecular tumor clocks. Nat. Med. 2:676–681.

60. Siegel, E. C., and V. Bryson. 1964. Selection of resistant strains of Escherichiacoli by antibiotics and antibacterial agents: role of normal and mutatorstrains, p. 629–634. Antimicrob. Agents Chemother. 1963.

61. Siegel, E. C., and V. Bryson. 1967. Mutator gene of Escherichia coli B. J.Bacteriol. 94:38–47.

62. Singer, M., T. A. Baker, G. Schnitzler, S. M. Deischel, M. Goel, W. Dove,K. J. Jaacks, A. D. Grossman, J. W. Erickson, and C. A. Gross. 1989. Acollection of strains containing genetically linked alternating antibiotic re-sistance elements for genetic mapping of Escherichia coli. Microbiol. Rev. 53:1–24.

63. Slupska, M. S., C. Baikalov, R. Lloyd, and J. H. Miller. 1995. MutatortRNAs are encoded by the Escherichia coli mutator genes mutA and mutC:a novel pathway for mutagenesis. Proc. Natl. Acad. Sci. USA 93:4380–4385.

64. Smith, K. C. 1992. Spontaneous mutagenesis: experimental, genetic, andother factors. Mutat. Res. 277:139–162.

65. Sniegowski, P. D., P. J. Gerrish, and R. E. Lenski. 1997. Evolution of highmutation rates in experimental populations of E. coli. Nature 387:703–705.

66. Strand, M., T. A. Prolla, R. M. Liskay, and T. D. Petes. 1993. Destabilizationof tracts of simple repetitive DNA in yeast by mutations affecting DNAmismatch repair. Nature 365:274–276.

67. Taddei, F., M. Radman, J. Maynard-Smith, B. Toupance, P. H. Gouyon, andB. Godelle. 1997. Role of mutator alleles in adaptive evolution. Nature 387:700–702.

68. Treffers, H. P., V. Spinelli, and N. O. Belser. 1954. A factor (or mutatorgene) influencing mutation rates in Escherichia coli. Proc. Natl. Acad. Sci.USA 40:1064–1071.

69. Umar, A., and T. A. Kunkel. 1996. DNA-replication fidelity, mismatch repairand genome instability in cancer cells. Eur. J. Biochem. 238:297–307.

70. Weinberg, R. A. 1996. How cancer arises: an explosion of research is uncov-ering the long-hidden molecular underpinnings of cancer—and suggestingnew therapies. Sci. Am. 275:62–70.



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

direct selection for mutators in escherichia coli · low spontaneous mutation rate might not...

Documents