Our approach to the isolation of DNA mismatch-correction-deficient mutants was based upon the isolation of 2-aminopurine-resistant second-site revertants of Escherichia coli dam -mutants. We isolated such second-site revertants which, when separated from the dam -mutation, have a mutator character of their own. These new mutators all mapped at three known mutator loci, mutH, mutL, and mutS, which exhibit the same mutagenic spectrum as the dam -mutator: increased levels of base substitution and frameshift mutations. The mutator potencies of double and triple mutmutants suggest that these mutators are involved in the same general mismatch-repair pathway. All these mutations result in a hyper-recombination phenotype, but in four-factor crosses among X phages, a specific loss of intragenic recombination (Pam3 X Pam8O) was found in mutL and mutS mutants, as would be predicted from the postulated role of mismatch correction in gene conversion and high negative interference phenomena.The existence of an excision-repair system acting upon mismatched base pairs in the DNA has been postulated in order to account for gene conversion (1, 2), high negative interference (3), and map expansion phenomena (4) (for review, see ref. 5). The possibility of an involvement of mismatch repair in the suppression of spontaneous mutations was indicated by the discovery that pneumococcus hex-and Escherichia coli uvrE mutants, which are probably deficient in the repair of some mismatched base pairs, appear to be spontaneous mutators as well (6, 7). Furthermore, mismatch repair has been implicated in the avoidance of mutagenesis by 5-bromouracil (5-BrUra) (8).However, the existence of a mismatch repair system to efficiently correct replication errors implies that a strand discrimination mechanism must exist ensuring that the excision of the mismatched base occurs exclusively from the newly synthesized strand. Because DNA methylation is a postreplicative process [i.e., newly synthesized strands are undermethylated (9)], it has been suggested that DNA methylation is one possible means of discrimination between old (methylated) and new (undermethylated) strands (10-12). This suggestion is supported by the observation that E. coli dam-mutants deficient in general methylation of adenine residues (13) occurring within the 5'G-A-T-C3' sequence (14) are also spontaneous mutators (15). Direct evidence in favor of the above hypothesis was obtained by using transfection assays with heteroduplex A DNA differing in the degree of methylation and carrying different genetic markers.tf Furthermore, a role for damdependent methylation in DNA strand discrimination in the elimination of the mutagenic effects of base analogs has been indicated by the sensitivity and hypermutability of E. coli dam-mutants for the base analogs 2-aminopurine (2APur) and 5-BrUra (11, 12).Our strategy to isolate mutants defective in adenine-methylation-instructed mismatch correction was based upon the sensitivity of the dam-mutants to 2APur (11). Fig. 1 Tables 1 and 3. Medi...

No abstract

No abstract

A model is presented for deletion mutations whose formation is mediated by palindromic and quasipalindromic DNA sequences. It proposes that the self-complementarity of palindromes allows the formation of DNA secondary structures that serve as deletion intermediates. The structures juxtapose the end points of the deletion and thus direct deletion specificity. While misaligned DNA intermediates that explain deletion termini occurring in repeated DNA sequences have been described, no explanations have been offered for deletion termini occurring in other sequences. The DNA secondary structures whose formation is mediated by palindromic sequences appear to explain many of these. In this paper, secondary-structure intermediates are described for a series of spontaneous deletions of known sequence in the lad gene of Escherichia coli. The model is supported by its failure to predict structures that can juxtapose simulated deletion termini in the lacI gene. We have found a strong association between palindromic sequences and repeated sequences at lacl deletion termini that suggests the joint participation of repeated and palindromic DNA sequences in the formation of some deletions. Sequences of deletions in other organisms also suggest the participation of palindromic DNA sequences in the formation of deletions.Local DNA sequences can strongly influence the frequency of mutation (1). The frequent association of deletion termini with repeated DNA sequences (2-4) suggests that deletion mutations are no exception. The In its simplest form, our model proposes that DNA secondary structures whose formation is potentiated by palindromic or quasipalindromic DNA sequences participate in deletion formation through the juxtaposition of deletion end points. The inherent self-complementarity of palindromic DNA sequences permits the formation of cruciform or hairpin structures in nucleic acids, precisely juxtaposing otherwise distant bases (6). Repeated DNA sequences are not required.The termini of the E. coli lacd deletion S86 cannot be juxtaposed by a misalignment involving repeated sequences. However, the deletion termini are located precisely at the ends of a quasipalindromic DNA sequence that includes the entire deletion. The formation of a DNA hairpin or cruciform structure in the wild-type DNA sequence places the deletion termini immediately adjacent to one another, rather than separated by their normal linear distance of 27 base pairs (Fig. 1A). This structural intermediate would produce the S86 deletion if it served as a substrate for excision or as a template for DNA synthesis (Fig. 1B). The latter process might be mediated by either DNA polymerase or DNA ligase. A ligation across such a hairpin stem might also be responsible for deletions if the hairpin served as a substrate for nucleases that removed aberrantly aligned regions from ordinary double-stranded DNA. Whatever the precise molecular mechanism(s) responsible for the production of deletions from such DNA structures, the prediction of the model is that the ...

No abstract