Conditional Mutator Gene in Escherichia coli: Isolation, Mapping, and Effector Studies

A mutator gene, mutD5, whose phenotype is conditional, has been identified in Escherichia coli. By P1 transduction it has been shown to lie at about 5.7 min on the chromosome, being co-transduced with proA and argF. In rich medium, streptomycin- and nalidixic acid-resistant mutation frequencies are 50 to 100 times higher than those in minimal medium. In minimal medium, the mutD5-induced mutation frequencies are still 50 to 100 times above co-isogenic wild-type (mut(+)) levels. Similar results were obtained with all markers tested. Mutant frequencies can be raised by thymidine in the medium at concentrations as low as 0.04 muM, or by the endogenous generation of thymidine from thymine plus a deoxyribosyl donor. Deoxyadenosine, various ribonucleosides, thymine, and 2-deoxyribose do not stimulate mutation. None of these effects are related to growth rate, since growth rate and mutation rate can be decoupled completely.     

The regulation of the ammonia assimilatory enzymes in Rel+ and Rel- strains of Salmonella typhimurium

Summary The influence of the relA1 mutation on the regulation of the ammonia assimilatory enzymes, glutamate dehydrogenase (EC 1.4.1.4), glutamine synthetase (EC 6.3.1.2), and glutamate synthase (EC 1.4.1.3), was examined. When cells grown in rich media (either Luria broth or glucose-ammonia plus casamino acids) were transferred to a glucose-ammonia medium, the relA mutant failed to resume growth and did not have the same increase in any of the assimilatory enzyme activities as the rel ⁺ strain. This effect was particularly dramatic for glutamate dehydrogenase, which increased 6-fold in the rel ⁺ strain. Measurements of the guanosine nucleotide concentrations showed that the rel ⁺ strain had a ppGpp concentration about 9 times that of the relA mutant 5 min after the shift to minimal medium. These results are consistent with those for other biosynthetic enzymes and show that the ammonia assimilatory enzymes require a relA product for their synthesis during shifts from rich to minimal media. In addition, we examined the response of these strains to a change in nitrogen source. The relA mutant again failed to resume growth after a shift from glucose-ammonia to glucose-arginine medium. Even though the ppGpp concentration did not increase, the rel ⁺ strain grew and increased glutamine synthetase activities about 2-fold. These changes in the absence of increased ppGpp levels suggest that some other relA-mediated function is important during this change in nitrogen source.     

Alteration of ribosomal protein L6 in mutants of Escherichia coli resistant to gentamicin.

Summary Spontaneous and ethylmethane-sulfonate induced mutants of Escherichia coli resistant to gentamicin sulfate were isolated and investigated for alterations in the ribosomal protein pattern. It was found by two-dimensional polyacrylamide gel electrophoresis that three independently isolated strains did not show any spot for ribosomal protein L6. On cochromatography of radioactively labelled mutant and wild-type ribosomal proteins on carboxymethyl-cellulose columns a shift of the elution position of protein L6 was observed, the new elution positions being characteristic for the individual mutants analyzed which indicates that they possess different alterations in the L6 primary structure.Genetic analysis showed that the gentamicin resistant strains contain at least two mutations. One of them correlates with the altered L6 protein and causes an increased minimal inhibitory concentration of the drug by about 5 to 10-fold. The other mutation is not yet biochemically characterized. Its presence is connected with an about 10 to 20-fold increase in the resistance. Both mutations, when put together, confer resistance to 50 to 100 g/ml of the antibiotic in a low salt rich medium and to 1 mg/ml in a defined medium with a high concentration of phosphate. Cross-resistance analysis demonstrated that the three gentamicin-resistant (double-mutant) strains with the altered L6 protein are resistant to 50–100 g per ml of all other aminoglycoside antibioties tested. This forms a sharp contrast to the streptomycin resistance mutations present in strA1, strA40 or strA60 mutants which do not confer markedly increased levels of resistance to most of the other aminoglycosides.     

Increased divalent metal ion uptake associated with hydrogenase constitutive phenotypes of Bradyrhizobium japonicum

Hydrogenase-constitutive (Hup^c) mutants of Bradyrhizobium japonicum were previously shown to accumulate more nickel than the wild-type strain. In a 2 h period Hup^c strains JH101 and JH103 also accumulated 2- to 3-fold more Mg²⁺, Zn²⁺ and Cu²⁺, and about 4-fold more Co²⁺ and Mn²⁺ than the wild-type strain JH. Init uptake rates (first 10 min) by the Hup^c strains were also greater for all the metals. The mutation in the Hup^c strains affecting a trans-acting regulator of the hup structural genes appears to have also amplified a metal uptake/accumulation process common to many divalent metal ions. From efflux experiments (suspension of cells in metal-free medium after metal accumulation) to determine the degree of dissociation of each metal with the cells it was concluded that Zn²⁺, like Ni²⁺, was rapidly and tightly cell-associated. In contrast, about 50% of the accumulated Cu²⁺ and about 30% of the Mn²⁺ was effluxed within 2 h by both the Hup^c and wild-type strains. Cobalt was more tightly cell-associated than Mn²⁺ or Cu²⁺, as the strains effluxed about 26% of the previously accumulated metal in 2 h. Even after accounting for effluxed metal, the Hup^c strains retained more of each metal than the wild-type. The increased metal accumulation by Hup^c strains could not be accounted for solely at the level of transport, as known metabolic inhibitors (carbonyl cyanide m-chlorophenylhydrazone and nigericin) of nickel transport partially inhibited (1 h) accumulation of only some (magnesium, zinc and copper) of the other metals. Hydrogenase-derepressed wild-type cells exhibited slightly higher (22–27% more) 2 h accumulation capacity for some of the metals (nickel, zinc and copper) than did non-derepressed cells, but not to the 2- to 4-fold greater level observed for Hup^c strains compared with the wild-type. The Hup^c strains JH101 and JH103 do not synthesize more capsular/cell wall carbohydrate than the wild-type strain.     

Growth and translation elongation rate are sensitive to the concentration of EF-Tu

We have used quantitative immunoblotting to estimate the amount of EF-Tu in a variety of S. typhimurium strains with wild-type, mutant, insertionally inactivated or plasmid-borne tuf genes. In the same strains we have measured translation elongation rate, exponential growth rate and the level of nonsense codon readthrough. In the wild-type strain, at moderate to fast growth rates, our data show that EF-Tu makes up 8–9% of total cell protein. Strains with either of the tuf genes insertionally inactivated have 65% of the wild-type EF-Tu level, irrespective of which tuf gene remains active, or whether that gene is wild-type or a kirromycin-resistant mutant. Strains with only one active tuf gene have reduced growth and translation elongation rates. From the magnitude of the reduction in elongation rate relative to the level of EF-Tu we calculate that in glucose minimal medium the in vivo saturation level of wild-type ribosomes by ternary complexes is only 63%. Strains with a ribosome mutation causing a poor interaction with ternary complex are non-viable on minimal medium when the level of EF-Tu is reduced.     

Stimulation of the multiplication of Micrococcus luteus by an autocrine growth factor

Viable cells of Micrococcus luteus secrete a proteineous growth factor (Rpf) which promotes the resuscitation of dormant, nongrowing cells to yield normal, colony-forming bacteria. When washed M. luteus cells were used as an inoculum, there was a pronounced influence of Rpf on the true lag phase and cell growth on lactate minimal medium. In the absence of Rpf, there was no increase in colony-forming units for up to 10 days. When the inoculum contained less than 10⁵ cells ml^–1, macroscopically observable M. luteus growth was not obtained in succinate minimal medium unless Rpf was added. Incubation of M. luteus in the stationary phase for 100 h resulted in a failure of the cells to grow in lactate minimal medium from inocula of small size although the viability of these cells was close to 100% as estimated using agar plates made from lactate minimal medium or rich medium. The underestimation of viable cells by the most-probable-number (MPN) method in comparsion with colony-forming units was equivalent to the requirement that at least 10⁵ cells grown on succinate medium, 10³ cells from old stationary phase, or approximately 10–500 washed cells are required per millilitre of inoculum for growth to lead to visible turbidity. The addition of Rpf in the MPN dilutions led to an increase of the viable cell numbers estimated to approximately the same levels as those determined by colony-forming units. Thus, a basic principle of microbiology –“one cell-one culture”– may not be applicable in some circumstances in which the metabolic activity of “starter” cells is not sufficient to produce enough autocrine growth factor to support cell multiplication. Received: 7 December 1998 / Accepted: 7 April 1999     

The effects of different thymidine concentrations on DNA replication in pea-root cells synchronized by a protracted 5-fluorodeoxyuridine treatment

Single-cell and DNA fiber autoradiography, cytophotometry and velocity sedimentation in alkaline sucrose gradients were used to analyse DNA replication and nascent replicon maturation in 5-fluorodeoxyuridine (FUdR)-synchronized cells of Pisum sativum. The replicon size was not significantly changed by the protracted FUdR treatment. When the synchronized cells were released from the inhibitor, labeled with [³H]TdR for 30 min, and chased in medium containing 1 × 10⁻⁶ M or lower concentrations of cold thymidine, DNA replication stopped after approx. 25% of the genome had replicated, and the nascent strands failed to grow above 9–12 × 10⁶ D single-stranded (ss) DNA. When the cells were chased in medium with 1 × 10⁻⁵ M cold thymidine, the DNA content of the labeled cells steadily increased with time and the size of the nascent molecules grew continuously until replicon size was achieved; then they were accumulated at replicon size until the cells arrived in late S or G2. When the FUdR-synchronized cells were chased in medium containing 1 × 10⁻⁴ M cold thymidine, the size of the nascent strands increased continuously with time, indicating that some neighbouring nascent replicons were joined as soon as they completed their replication. These observations led us to postulate that in FUdR-synchronized cells the rates of chain elongation, cell progression through the S phase and nascent replicon maturation are controlled by thymidine availability.     

Transductional construction of aspartase-hyperproducing strain of Escherichia coli B

Two aspartase-overproducing mutants of Escherichia coli B were characterized. Strain EAPc7 had a mutation enhancing aspartase formation in the region of aspartase gene. This mutation did not affect catabolite repression by aspartase. Strain EAPc244 showed a high cAMP content and an elevated adenylate cyclase activity. This mutation was closely linked to the ilv locus and caused the release of catabolite repression for various catabolite repression-sensitive enzymes, resulting in overproduction of adenylate cyclase. This mutation was transduced to an Ile⁻ strain derived from strain EAPc7 using the Ile⁺ selective marker. The constructed strain AT202, having the above 2 mutations, produced about 3-fold and 18-fold more aspartase than did the 2 parent strains and the wild-type strain, respectively, when cultured in the medium used for industrial production of aspartase. Strain AT202 maintained stably high aspartase activity after 30 cell generations. On the other hand, in E. coli K-12 harboring the aspA⁺ recombinant plasmid pYT471 (pBR322-aspA⁺), the activity decreased to the E. coli K-12 level. Hence, strain AT202 is more advantageous for industrial production of l-aspartic acid than cells harboring the aspA⁺-recombinant plasmid pYT471.     

Expression of Vitreoscilla hemoglobin in Gordonia amarae enhances biosurfactant production

The gene (vgb) encoding Vitreoscilla (bacterial) hemoglobin (VHb) was electroporated into Gordonia amarae, where it was stably maintained, and expressed at about 4 nmol VHb g⁻¹ of cells. The maximum cell mass (OD₆₀₀) of vgb-bearing G. amarae was greater than that of untransformed G. amarae for a variety of media and aeration conditions (2.8-fold under normal aeration and 3.4-fold under limited aeration in rich medium, and 3.5-fold under normal aeration and 3.2-fold under limited aeration in mineral salts medium). The maximum level of trehalose lipid from cultures grown in rich medium plus hexadecane was also increased for the recombinant strain, by 4.0-fold in broth and 1.8-fold in cells under normal aeration and 2.1-fold in broth and 1.4-fold in cells under limited aeration. Maximum overall biosurfactant production was also increased in the engineered strain, by 1.4-fold and 2.4-fold for limited and normal aeration, respectively. The engineered strain may be an improved source for producing purified biosurfactant or an aid to microorganisms bioremediating sparingly soluble contaminants in situ.     

A new class of p-fluorophenylalanine-resistant mutants in Neurospora crassa

A p-fluorophenylalanine-resistant mutant (acc ^phe ) which grows on minimal medium has an altered prephenate dehydrogenase and maps at the try-1 locus. Two other tyr-1 mutants which require tyrosine for normal growth can eventually grow on minimal or minimal plus p-fluorophenylalanine (FPA). The three different tyr-1 mutants all accumulate phenylalanine when incubated in minimal medium. FPA is incorporated into protein at only 10–15% the wild-type rate when mutant conidia are incubated in a minimal salts-glucose system. Under the same conditions, phenylalanine incorporation in the mutants is initially the same as in wild type. When tyrosine is included in the medium, resistance to FPA is lost, phenylalanine accumulation is prevented, and FPA is incorporated into protein at the wild-type rate. Tyrosine apparently prevents the overproduction of phenylalanine by preventing the overproduction of chorismate and prephenate.This work was supported, in part, by an Atomic Energy Commission grant to the Institute of Molecular Biophysics, the Florida State University, and by the Genetics Training Grant, funded by the National Institute of Health. It contains, in part, data from the doctoral thesis of the senior author, who was supported by a Florida State University Nuclear Fellowship and by a Public Health Service Fellowship.     

Generation of dormant forms by <Emphasis Type="Italic">Mycobacterium smegmatis</Emphasis> in the poststationary phase during gradual acidification of the medium

Persistent forms of the wild-type strain of Mycobacterium smegmatis and its mutants with inactivated devR and hlp genes were investigated. devR encodes the regulatory protein responsible for the formation of nonreplicating mycobacterial forms under hypoxia, and hlp codes for a histone-like protein. It has been found that a gradual decrease of pH in M. smegmatis wild-type poststationary cultures resulted in the formation of a special type of persisters. They significantly differed from vegetative cells in their properties and were represented by shortened ovoid forms with thickened cell walls. According to atomic force microscopy data, the size of the ovoid forms and vegetative cells was 1.2 × 0.9 μm and 3.7 × 0.8 μm, respectively. The metabolism level was markedly decreased in ovoid cells: the incorporation of [5,6-³H]uracil and thymidine was decreased 200- and 50-fold, respectively. The intracellular ATP content was lowered threefold. The ovoid forms that emerged in poststationary cultures in Sauton’s medium when the medium pH value was gradually decreased to 6.0 retained for a long time (9 months) the capacity to resume growth on rich solid and liquid medium. Compared to vegetative cells, the ovoid forms exhibited an elevated resistance to heating (60–80°C) and antibiotics (hygromycin, kanamycin, and tetracycline). The ovoid forms of the M. smegmatis wild-type strain were classified as dormant forms based on their survival capacity, resistance to deleterious factors, and structural peculiarities. The ovoid forms generated in poststationary cultures upon decreasing the pH value to 5.0 or below lost the colony-forming capacity. It was established that the capacity to form ovoid cells upon gradual decrease in the pH value to 6.0 was reduced in ΔdevR and hlp-0 mutants compared to the wild-type strain (generation of 5–6 and 40% dormant forms, respectively) The amount of M. smegmatis dormant cells formed correlated with the acidification degree of the medium. The model developed can be used in tests of new antibacterial preparations that effectively inhibit resuscitating mycobacterial dormant forms that persist in the host organism.     

Yersinia intermedia: A new species of enterobacteriaceae composed of rhamnose-positive,melibiose-positive,raffinose-positive strains (formerly calledYersinia enterocolitica orYersinia enterocolitica-like)

The drugs griseofulvin (10 μg/ml), nalidixic acid (0.05 μg/ml), quinine dihydrochloride (50 μg/ml), quinine ethylcarbonate (50 μg/ml), quinine urea hydrochloride (50 μg/ml), quinine lactate (50 μg/ml), and pamaquine (50 μg/ml) were chosen for laboratory studies. The minimal inhibitory concentration of the drug was used for determining the range of drug concentration needed to produce “mutational synergism” with ultraviolet radiation. Forward mutation from streptomycin sensitivity to resistance was used as a marker for mutagenicity. No stimulatory or inhibitory effects were noted on viable counts and mutation frequency, when the drugs were added (20–60 μg/ml) to the growth medium of unirradiatedEscherichia coli HCR⁺, HCR⁻, and irradiated HCR⁻ strains. These drugs increased mutation frequency and lethality of irradiated HCR⁺ bacteria. Incorporation of adenine (6 μm) into the minimal expression medium reverses the mutagenic effect of chloroquine. Chloroquine (50 μg/ml) did not interfere with the photoactivation of irradiated HCR⁺ cells. Our findings suggest that these chemicals selectively interfere with excision-repair.     

SeqA limits DnaA activity in replication from oriC in Escherichia coli

A mutant Escherichia coli that transforms minichromosomes with high efficiency in the absence of Dam methylation has been Isolated and the mutation mapped to 16.25 min on the E. coli map. The mutant strain containing seqA2 is defective for growth in rich medium but not in minimal medium. A similar mutation In this gene, named seqA1, has also been isolated. Here we show that the product of the seqA gene, SeqA, normally acts as an inhibitor of chromosomal initiation. In the seqA2-containing mutant, the frequency of initiation increases by a factor of three. Introduction of the wild-type seqA gene on a low-copy plasmid suppresses the cold sensitivity of a dnaAcos mutant known to overinitiate at temperatures below 39°C. In addition, the seqA2 mutation is a suppressor of several dnaA (Ts) alleles. The seqA2 mutant overinitiates replication from oriC and displays the asynchronous initiation phenotype. Also the seqA2 mutant has an elevated level of DnaA protein (twofold). The introduction of minichromosomes or a low-copy-number plasmid carrying five DnaA-boxes from the oriC region increases the growth rate of the seqA2 mutant in rich medium to the wild-type level, reduces overinitiation but does not restore synchrony. We propose that the role of SeqA is to limit the activity level of the E. coli regulator of chromosome initiation, DnaA.     

Origin of thymidine kinase deficient (TK-) haploid frog cells via an intermediate thymidine transport deficient (TT-) phenotype

Wild-type cultured cells of the frog cell line ICR 2A give rise to 5-bromodeoxyridine (BUdR)-resistant colonies only when the selecting concentration of the drug is 5 × 10^?5 M or lower. The progeny of these colonies multiply in 10^?4 M BUdR; resistance is correlated with the absence of a thymidine (TdR)-specific transport reaction with a K^m in the range of 2–7 × 10^?4 M. All of the TdR transport-deficient (TT-) isolates examined (25) had TdR kinase activity (4% to 100% of wild-type). Variants deficient in TdR kinase activity (5% of wild-type) were obtained by exposing TT-cultures to 10^?3 M BUdR. The TK - variants multply continuously in 10^?3 M BUdR and retain the phenotype after prolonged culture in the absence of the drug. The frequency with which they occur is increased 20 to 50 fold by prior treatment of the culture with ICR 191, an acridine mustard mutagen. In haploid cells, it would be expected that TK- variants would arise in equal numbers from wild-type and TT- cultures if loss TdR kinase occurred independently of loss of the transport reaction. However, wild-type cells give no colonies resistant to 10^?3 M BUdR under conditions the give 1 to 50 colonies per million TT- cells. The TT- phenotype seems to be a required intermediate state in the origin of the TK- phenotype. Therefore, the TK- clones described above are unlikely to be products of mutation at a single genetic locus.     

A Requirement for DNA Synthesis during Auxin Induction of Cell Enlargement in Tobacco Pith Tissue

IAA (indoleacetic acid) is known to induce cell enlargement without cell division in tobacco pith explants grown on an agar medium without added cytokinin. The very long lag period before IAA (2 × 10^?5M) stimulates growth, about 3 days, can be useful to study the metabolic changes which lead to the promotion of growth. When the disks are transferred to a medium without IAA after 2 days or less of treatment with IAA, the IAA does not stimulate growth. Disks transferred after 3 days, subsequently show an auxin response, almost as great as those given IAA continuously. At 5 × 10^?4M, 5-fluorodeoxyuridine (FUDR), which inhibits DNA synthesis by blocking formation of thymidylate, completely suppresses the lAA-induced growth if it is added together with the IAA or 1 day later. When the FUDR is given 2 days after the IAA, there is a small increment of auxin-induced growth, and an even greater amount if added after 3 days. The period when exogenous auxin must be present to stimulate growth corresponds to the period of FUDR sensitivity. The FUDR inhibition is prevented by thymidine but not by uridine. Other inhibitors of DNA synthesis, hydroxyurea and fluorouracil, also inhibit auxin-induced growth. Thus DNA synthesis seems to be required for auxin induction of cell enlargement in tobacco pith explants. In contrast, FUDR does not inhibit auxin-induced growth in corn coleoptile and artichoke tuber sections.     

Utilization of the label from bacterial [3H] DNA by isolated barley roots and embryos under different conditions

[³H] DNA fromEscherichia coli and [³H] thymidine were applied, in sterile conditions, on isolated barley embryos and on roots excised from these embryos, both cultivated in the liquid medium and on halves of barley seeds, through the endosperm bridge. In embryos and roots, the labelled compounds were applied in 1.5% sucrose + 0.2 SSC alone, or together with either unlabelled thymidine or DEAE-dextran. Similar labelling indices were found after [³H] thymidine and [³H] DNA treatment which shows that the activity of [³H] DNA is utilized during the S phase. After application of [³H] thymidine, only cell nuclei in S phase were labelled. After the application of [³H] DNA an extranuclear label, in addition to the labelling of nuclei in the S phase, was observed in some experimental variants. The density of label above labelled nuclei after [³H] DNA treatment sharply decreased when unlabelled thymidine or DEAE-dextran was added, while the density of label above nuclei labelled by [³H] thymidine decreased when unlabelled thymidine but not DEAE-dextran was added. The labelling of nuclei with the label from [³H] DNA is the result of degradation of exogenous DNA reutilization of low molecular weight products. Extranuclear labelling is most probably due to the polymerous or partly degraded DNA.     

Cycloheximide resistance in Chinese hamster cells I. Spontaneous mutagenesis

Resistance to cycloheximide (CHM) was studied in cultured Chinese hamster cells. Concentrations of CHM above 5.10⁻⁷M were toxic for the cells. At concentrations above 9.10⁻⁷M no colonies were recovered in selective medium. 15 resistant clones of independent origin were isolated in selective medium containing 7.10⁻⁷M CHM. Resistance was stable when the cells were cultured under non-selective conditions. The spontaneously mutation rate was determined by the fluctuation test. Mutations to CHM resistance arose spontaneously. The spontaneous mutation rate to CHM resistance was about 10⁻⁵.     

Mutator suppression and escape from replication error-induced extinction in yeast

Cells rely on a network of conserved pathways to govern DNA replication fidelity. Loss of polymerase proofreading or mismatch repair elevates spontaneous mutation and facilitates cellular adaptation. However, double mutants are inviable, suggesting that extreme mutation rates exceed an error threshold. Here we combine alleles that affect DNA polymerase δ (Pol δ) proofreading and mismatch repair to define the maximal error rate in haploid yeast and to characterize genetic suppressors of mutator phenotypes. We show that populations tolerate mutation rates 1,000-fold above wild-type levels but collapse when the rate exceeds 10⁻³ inactivating mutations per gene per cell division. Variants that escape this error-induced extinction (eex) rapidly emerge from mutator clones. One-third of the escape mutants result from second-site changes in Pol δ that suppress the proofreading-deficient phenotype, while two-thirds are extragenic. The structural locations of the Pol δ changes suggest multiple antimutator mechanisms. Our studies reveal the transient nature of eukaryotic mutators and show that mutator phenotypes are readily suppressed by genetic adaptation. This has implications for the role of mutator phenotypes in cancer.     

Properties of the nonlethal recombinational repair deficient mutants of bacteriophage T4

Summary The rate at which ³H thymidine is incorporated into DNA is increased in T4w-infected cells compared to wild-type when measured late in infection under conditions of low thymidine concentration. This increased DNA synthesis is sensitive to hydroxyurea but not to mitomycin C, and can be prevented by the addition of chloramphenicol early in infection. Also, DNA replicative intermediates isolated from T4w-infected cells late in infection sediment significantly faster than those isolated from wild-type-infected cells. In contrast, DNA replicative intermediates isolated from T4x-or T4y-infected cells sediment more slowly than those produced by wild-type T4. Cells coinfected with wild-type T4⁺ and T4x, y or w; or T4w and T4x or y, produce wild-type DNA replicative intermediates. Cells coinfected with T4x and T4y produce more slowly sedimenting DNA replicative intermediates. Cells coinfected with T4w and wild-type T4 show wild-type rates of DNA synthesis while cells coinfected with T4w and T4x or T4y show increased rates of DNA synthesis over that observed with wild-type alone.     

Strand targeting signal(s) for in vivo mutation avoidance by post-replication mismatch repair in Escherichia coli

Summary The involvement of GATC sites in directing mismatch correction for the elimination of replication errors in Escherichia coli was investigated in vivo by analyzing mutation rates for a gene carried on a series of related plasmids that contain 2, 1 and 0 such sites. This gene encoding chloramphenicol acetyl transferase (Cat protein) was inactivated by a point mutation. In vivo mutations restoring resistance to chloramphenicol were scored in mismatch repair proficient (mut ⁺) and deficient (mutHLS^-) strains. In mut ⁺ cells, reduction of GATC sites from 2 to 0 increased mutation rates approximately 10-fold. Removal of the GATC site distal to the cat ^- mutation increased the rate of mutation less than 2-fold, indicating that mismatch repair can proceed normally with a single site. The mutation rate increased 3-fold after removal of the GATC site proximal to the mutation. In the absence of a GATC site, mutL^- and mutS^- strains exhibited a 2- to 3-fold increased mutation rate as compared to isogenic mutH^- and mut ⁺ strains. This indicates that 50%–70% of replication errors can be corrected in a mutLS-dependent way in the absence of any GATC site to target mismatch correction to newly synthesized DNA strands. Other strand targeting signals, possibly single strand discontinuities, might be used in mutLS-dependent repair