Palindromy and the Location of Deletion Endpoints in Escherichia Coli

The contributions of direct and inverted repeats to deletion formation were studied by characterizing Ampr revertants of plasmids with a series of insertion mutations at a specific site in the pBR322 ampicillin resistance (amp) gene. The inserts at this site are palindromic, variable in length, and bracketed by 9- or 10-bp direct repeats of amp sequence. There is an additional direct repeat composed of 4 bp within the insert and 4 bp of adjoining amp sequence. DNA sequencing and colony hybridization of Ampr revertants showed that they contained either the parental amp sequence, implying deletion endpoints in the flanking 9- or 10-bp repeats, or a specific 1-bp substitution, implying endpoints in the 4-bp repeats. Although generally direct repeats seem to be used as deletion endpoints with a frequency proportional to their lengths, we found that with uninterrupted palindromes longer than 32 bp, the majority of deletions ended in the 4 bp, not the 9- or 10-bp repeats. This preferential use of the shorter direct repeats associated with palindromes is interpreted according to a DNA synthesis-error model in which hairpin structures formed by intrastrand pairing foster the slippage of nascent strands during DNA synthesis.     

Deletions in Plasmid Pbr322: Replication Slippage Involving Leading and Lagging Strands

We test here whether a class of deletions likely to result from errors during DNA replication arise preferentially during synthesis of either the leading or the lagging DNA strand. Deletions were obtained by reversion of particular insertion mutant alleles of the pBR322 amp gene. The alleles contain insertions of palindromic DNAs bracketed by 9-bp direct repeats of amp sequence; in addition, bp 2 to 5 in one arm of the palindrome form a direct repeat with 4 bp of adjoining amp sequence. Prior work had shown that reversion to Ampr results from deletions with endpoints in the 8- or 4-bp repeat, and that the 4-bp repeats are used preferentially because one of them is in the palindrome. To test the role of leading and lagging strand synthesis in deletion formation, we reversed the direction of replication of the amp gene by inverting the pBR322 replication origin, and also constructed new mutant alleles with a 4-bp repeat starting counterclockwise rather than clockwise of the insertion. In both cases the 4-bp repeats were used preferentially as deletion endpoints. A model is presented in which deletions arise during elongation of the strand that copies the palindrome before the adjoining 4-bp repeat, and in which preferential use of the 4-bp repeats independent of the overall direction of replication implies that deletions arise during syntheses of both leading and lagging strands.     

The effect of the length of direct repeats and the presence of palindromes on deletion between directly repeated DNA sequences in bacteriophage T7.

The frequency of genetic deletion between directly repeated DNA sequences in bacteriophage T7 was measured as a function of the length of the direct repeat. The non-essential ligase gene (gene 1.3) of bacteriophage T7 was interrupted with pieces of synthetic DNA bracketed by direct repeats of various lengths. Deletion of these 76 bp long inserts was too low to be measured when the direct repeats were less than 6 bp long. However, the frequency of deletion of inserts with longer direct repeats increased exponentially as the length of the repeats increased from 8 to 20 bp. When inverted repeats (palindromes) were designed in the midst of the insert there was essentially no increase in deletion frequency between 10 bp direct repeats. But, the same palindromic sequences increased the deletion frequency between 5 bp direct repeats by at least two orders of magnitude. Thus, in this system homology at the endpoints is a more important determinant of deletion frequency than is the presence of palindromes between the direct repeats.     

Deletion mutagenesis independent of recombination in bacteriophage T7.

Deletion between directly repeated DNA sequences in bacteriophage T7-infected Escherichia coli was examined. The phage ligase gene was interrupted by insertion of synthetic DNA designed so that the inserts were bracketed by 10-bp direct repeats. Deletion between the direct repeats eliminated the insert and restored the ability of the phage to make its own ligase. The deletion frequency of inserts of 85 bp or less was of the order of 10(-6) deletions per replication. The deletion frequency dropped sharply in the range between 85 and 94 bp and then decreased at a much lower rate over the range from 94 to 900 bp. To see whether a deletion was predominantly caused by intermolecular recombination between the leftmost direct repeat on one chromosome and the rightmost direct repeat on a distinct chromosome, genetic markers were introduced to the left and right of the insert in the ligase gene. Short deletions of 29 bp and longer deletions of approximately 350 bp were examined in this way. Phage which underwent deletion between the direct repeats had the same frequency of recombination between the left and right flanking markers as was found in controls in which no deletion events took place. These data argue against intermolecular recombination between direct repeats as a major factor in deletion in T7-infected E. coli.     

The Influence of Primary and Secondary DNA Structure in Deletion and Duplication between Direct Repeats in Escherichia Coli

We describe a system to measure the frequency of both deletions and duplications between direct repeats. Short 17- and 18-bp palindromic and nonpalindromic DNA sequences were cloned into the EcoRI site within the chloramphenicol acetyltransferase gene of plasmids pBR325 and pJT7. This creates an insert between direct repeated EcoRI sites and results in a chloramphenicol-sensitive phenotype. Selection for chloramphenicol resistance was utilized to select chloramphenicol resistant revertants that included those with precise deletion of the insert from plasmid pBR325 and duplication of the insert in plasmid pJT7. The frequency of deletion or duplication varied more than 500-fold depending on the sequence of the short sequence inserted into the EcoRI site. For the nonpalindromic inserts, multiple internal direct repeats and the length of the direct repeats appear to influence the frequency of deletion. Certain palindromic DNA sequences with the potential to form DNA hairpin structures that might stabilize the misalignment of direct repeats had a high frequency of deletion. Other DNA sequences with the potential to form structures that might destabilize misalignment of direct repeats had a very low frequency of deletion. Duplication mutations occurred at the highest frequency when the DNA between the direct repeats contained no direct or inverted repeats. The presence of inverted repeats dramatically reduced the frequency of duplications. The results support the slippage-misalignment model, suggesting that misalignment occurring during DNA replication leads to deletion and duplication mutations. The results also support the idea that the formation of DNA secondary structures during DNA replication can facilitate and direct specific mutagenic events.     

Local DNA Sequence Control of Deletion Formation in Escherichia coli Plasmid Pbr322

The specificity of deletion formation was studied using tests involving reversion of palindromic insertion mutations. Insertions of a Tn5-related transposon at 13 sites in the ampicillin-resistance (amp) gene of plasmid pBR322 were shortened to a nested set of perfect palindromes, 22, 32 and 90 bp long. We monitored frequencies of reversion to Ampr, which is the result of deletion of the palindrome plus one copy of the flanking 9 bp direct repeats (which had been formed by transposition). Revertant frequencies were found to depend on the location and the sequence of the palindromic insert. Changing a 45-kb interrupted palindrome to a 22-bp perfect palindrome stimulated deletion formation by factors of from fourfold to 545-fold among the 13 sites, while elongation of the perfect palindrome from 22 to 90 bp stimulated deletion formation by factors of from eight- to 18,000-fold. We conclude that deletion formation is strongly affected by subtle features of DNA sequence or conformation, both inside and outside the deleted segment, and that these effects may reflect specific interactions of DNA processing proteins with template DNAs.     

RecA-independent high-frequency deletion of recombinant cosmid DNA in Escherichia coli

Segments of DNA were deleted from recombinant cosmid DNAs during propagation in Escherichia coli hosts in liquid culture. DNAs of more than 1000 cosmids propagated in various E. coli hosts were analysed by agarose gel electrophoresis (AGE). The effects of vectors, insert DNAs and host genetic characters on the formation of deletions were examined. The probability of deletion and the pattern of deletion bands observed by AGE differed from clone to clone, and after extensive culture the deletion band patterns remained almost constant during further culture. Most recombinant clones eventually showed deletion during prolonged liquid culture. Mutations in the recA gene of E. coli hosts, including a deletion mutation, did not prevent deletion. Most deletions occurred in the insert portions of cosmid DNAs. Nucleotide sequence analysis of six deletion junctions in test cosmid cMB15 demonstrated that deletions occurred between two short complete direct repeats of about 4-10 bp, irrespective of whether the cosmid was propagated in a recA host or a rec+ host. Some deletions occurred at the same sites either in a recA host or a rec+ host. These results suggest that the deletion events are mainly mediated by a recA-independent recombination system(s) of E. coli host cells.     

Specificity of deletion events in pBR322

The reversion of mutations due to inserts of identical palindromic DNAs just 1-bp apart in the amp gene of plasmid pBR322 varied up to 3000-fold (U. DasGupta, K. Weston-Hafer, and D.E. Berg (1987) Genetics 115, 41-49). The experiments reported here show that the intrinsic frequencies of deletion from these sites are truly very different. Deletions were selected by the joint loss of sacB (sucrose sensitivity) and lacZ alpa genes cloned together at these sites, without requiring restoration of the ampr allele. We found that greater than 90% of deletions at each of these sites do restore the ampr allele. This result reinforces the view that the probability of forming a particular deletion depends strongly on the DNA sequence at its prospective endpoints.     

Analysis of repeat-mediated deletions in the mitochondrial genome of Saccharomyces cerevisiae

Mitochondrial DNA deletions and point mutations accumulate in an age-dependent manner in mammals. The mitochondrial genome in aging humans often displays a 4977-bp deletion flanked by short direct repeats. Additionally, direct repeats flank two-thirds of the reported mitochondrial DNA deletions. The mechanism by which these deletions arise is unknown, but direct-repeat-mediated deletions involving polymerase slippage, homologous recombination, and nonhomologous end joining have been proposed. We have developed a genetic reporter to measure the rate at which direct-repeat-mediated deletions arise in the mitochondrial genome of Saccharomyces cerevisiae. Here we analyze the effect of repeat size and heterology between repeats on the rate of deletions. We find that the dependence on homology for repeat-mediated deletions is linear down to 33 bp. Heterology between repeats does not affect the deletion rate substantially. Analysis of recombination products suggests that the deletions are produced by at least two different pathways, one that generates only deletions and one that appears to generate both deletions and reciprocal products of recombination. We discuss how this reporter may be used to identify the proteins in yeast that have an impact on the generation of direct-repeat-mediated deletions.     

On the Deletion of Inverted Repeated DNA in Escherichia Coli: Effects of Length, Thermal Stability, and Cruciform Formation in Vivo

We have studied the deletion of inverted repeats cloned into the EcoRI site within the CAT gene of plasmid pBR325. A cloned inverted repeat constitutes a palindrome that includes both EcoRI sites flanking the insert. In addition, the two EcoRI sites represent direct repeats flanking a region of palindromic symmetry. A current model for deletion between direct repeats involves the formation of DNA secondary structure which may stabilize the misalignment between the direct repeats during DNA replication. Our results are consistent with this model. We have analyzed deletion frequencies for several series of inverted repeats, ranging from 42 to 106 bp, that were designed to form cruciforms at low temperatures and at low superhelical densities. We demonstrate that length, thermal stability of base pairing in the hairpin stem, and ease of cruciform formation affect the frequency of deletion. In general, longer palindromes are less stable than shorter ones. The deletion frequency may be dependent on the thermal stability of base pairing involving approximately 16-20 bp from the base of the hairpin stem. The formation of cruciforms in vivo leads to a significant increase in the deletion frequency. A kinetic model is presented to describe the relationship between the physical-chemical properties of DNA structure and the deletion of inverted repeats in living cells.     

Deletion between direct repeats in T7 DNA stimulated by double-strand breaks.

An in vitro system based on extracts of Escherichia coli infected with bacteriophage T7 was used to study genetic deletions between directly repeated sequences. The frequency of deletion was highest under conditions in which the DNA was actively replicating. Deletion frequency increased markedly with the length of the direct repeat both in vitro and in vivo. When a T7 gene was interrupted by 93 bp of nonsense sequence flanked by 20-bp direct repeats, the region between the repeats was deleted in about 1 out of every 1,600 genomes during each round of replication. Very similar values were found for deletion frequency in vivo and in vitro. The deletion frequency was essentially unaffected by a recA mutation in the host. When a double-strand break was placed between the repeats, repair of this strand break was often accompanied by the deletion of the DNA between the direct repeats, suggesting that break rejoining could contribute to deletion during in vitro DNA replication.     

The E35 stopper mutant of Neurospora crassa: precise localization of deletion endpoints in mitochondrial DNA and evidence that the deleted DNA codes for a subunit of NADH dehydrogenase.

Two types of defective mitochondrial DNA molecules with large deletions (5 kbp and 40 kbp) have previously been identified in the stopper mutant, E35, of Neurospora crassa. The junction fragments spanning the deletion endpoints have now been cloned and sequenced, and their sequences compared with those of the corresponding wild-type fragments. We show that both types of defective mitochondrial DNAs result from deletions of sequences flanked by short direct repeats, which are themselves parts of larger inverted repeat sequences. In every case, the short direct repeat sequences consist of a run of pyrimidines in one strand and purines in the other. We also report the sequence of a 2151-bp HindIII fragment, which is deleted in both of the defective mitochondrial DNAs. Besides the previously identified gene for a methionine tRNA, the 2151-bp DNA sequence contains an open reading frame with the potential to code for a hydrophobic protein 583 amino acids long. This hydrophobic protein has three blocks of significant homology with proteins coded by URF2 found in other mitochondrial genomes. Since the mammalian mitochondrial URF2 has recently been shown to code for a subunit of NADH dehydrogenase, part of the DNA sequence missing in the E35 stopper mutant of N. crassa may also code for a subunit of NADH dehydrogenase.     

Precise and nearly-precise excision of the symmetrical inverted repeats of Tn5; common features of recA-independent deletion events in Escherichia coli

The transposon Tn5 contains a unique central region bordered by 1.5-kb inverted repeats. The in vitro deletion of the centre of Tn5, with a restriction endonuclease (XhoI) which cuts within the inverted repeats leads to the production of a palindrome on subsequent ligation. This palindromic region is unstable on subsequent transformation into Escherichia coli (Collins, 1981). Precise excision of the Tn5 region plus one copy of the bracketing 9-bp direct repeat occurred in about one-third of the transformants. The rest of the transformants contain only remnants of the inverted repeat. Sequence analysis indicated that deletion had occurred between short direct repeats. The precise excision of these "nearly precise" excision products continued with high frequency and was found to be affected by mutations that interfere with the normal precise excision of transposons. In a recB, sbcB host precise excision was markedly reduced. A common mechanism is proposed for all recA-independent deletions occurring in E. coli.     

Asymmetric crossing over in the spontaneous formation of large deletions in the tonB-trp region of the Escherichia coli K-12 chromosome

The chromosomal tonB gene of Escherichia coli was used as a target for the detection of spontaneous deletion mutations. The deletions were isolated in both recA ⁺ and recA ^? cells, and mutants carrying large deletions were identified because they also lacked part or all of the trp operon. The frequencies of tonB-trp deletion were 1.79?×?10^?9 and 1.09?×?10^?9 for recA ⁺ and recA ^? cells, respectively. We analyzed 12 deletions from recA ⁺ and 10 from recA ^? cells by cloning and direct sequencing. The deletions ranged in size from 5612?bp to 15142?bp for recA ⁺ and from 5428?bp to 13289 for recA ^? cells. Three deletions from recA ⁺ cells and five deletions from recA ^? cells were found to have occurred between short sequence repeats at the termini of the deletion, leaving one copy of the repeat in the mutant sequence. Seven deletions from recA ⁺ cells and three deletions from recA ^? cells did not have repeats at their termini; in these cases, the DNA sequences that are adjacent to the deletion termini in the wild-type are characterized by short (2–4?bp) repeats. From these results, a model is presented for the generation of deletion mutations which involves formation of an asymmetric crossover mediated by repeated sequences of 2- to 4-bp.     

Targeting DNA polymerase to DNA double-strand breaks reduces DNA deletion size and increases templated insertions generated by CRISPR/Cas9

Most insertions or deletions generated by CRISPR/Cas9 (clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9) endonucleases are short (<25 bp), but unpredictable on-target long DNA deletions (>500 bp) can be observed. The possibility of generating long on-target DNA deletions poses safety risks to somatic genome editing and makes the outcomes of genome editing less predictable. Methods for generating refined mutations are desirable but currently unavailable. Here, we show that fusing Escherichia coli DNA polymerase I or the Klenow fragment to Cas9 greatly increases the frequencies of 1-bp deletions and decreases >1-bp deletions or insertions. Importantly, doing so also greatly decreases the generation of long deletions, including those >2 kb. In addition, templated insertions (the insertion of the nucleotide 4 nt upstream of the protospacer adjacent motif) were increased relative to other insertions. Counteracting DNA resection was one of the mechanisms perturbing deletion sizes. Targeting DNA polymerase to double-strand breaks did not increase off-targets or base substitution rates around the cleavage sites, yet increased editing efficiency in primary cells. Our strategy makes it possible to generate refined DNA mutations for improved safety without sacrificing efficiency of genome editing.     

Molecular characteristics of spontaneous deletions in the hyperthermophilic archaeon Sulfolobus acidocaldarius

Prokaryotic genomes acquire and eliminate blocks of DNA sequence by lateral gene transfer and spontaneous deletion, respectively. The basic parameters of spontaneous deletion, which are expected to influence the course of genome evolution, have not been determined for any hyperthermophilic archaeon. We therefore screened a number of independent pyrimidine auxotrophs of Sulfolobus acidocaldarius for deletions and sequenced those detected. Deletions accounted for only 0.4% of spontaneous pyrE mutations, corresponding to a frequency of about 10(-8) per cell. Nucleotide sequence analysis of five independent deletions showed no significant association of the endpoints with short direct repeats, despite the fact that several such repeats occur within the pyrE gene and that duplication mutations in pyrE reverted at high frequencies. Endpoints of the spontaneous deletions did not coincide with short inverted repeats or potential stem-loop structures. No consensus sequence common to all the deletions could be identified, although two deletions showed the potential of being stabilized by octanucleotide sequences elsewhere in pyrE, and another pair of deletions shared an octanucleotide at their 3' ends. The unusually low frequency and low sequence dependence of spontaneous deletions in the S. acidocaldarius pyrE gene compared to other genetic systems could not be explained in terms of possible constraints imposed by the 5-fluoroorotate selection.     

Cloning, nucleotide sequence, and expression of the Brucella melitensis bp26 gene coding for a protein immunogenic in infected sheep

Abstract We have previously identified a Brucella melitensis 28 kDa cytosoluble protein (CP28) which was highly immunogenic in infected sheep and which in addition made possible the serological differentiation between infected and B. melitensis Rev.l vaccinated sheep. Monoclonal antibodies against CP28 were used to screen a B. melitensis 16M genomic library and to clone the corresponding gene. DNA sequencing of the gene encoding CP28 of B. melitensis 16M revealed that it was nearly identical to that of the recently published bp26 gene of Brucella abortus vaccine strain S19 coding for a periplasmic protein. The differences between the B. melitensis 16M gene and that of B. abortus S19 consisted of single nucleotide substitutions, one or two codon deletions, one codon addition, and most importantly a 21-bp deletion. The corresponding region of B. abortus S19 contains two 10-bp direct repeats which could have been involved in the genesis of the deletion. Expression of the B. melitensis 16M bp26 gene in Escherichia coli studied by the use of the monoclonal antibodies showed the same characteristics as reported for the B. abortus S19 bp26 gene, i.e. the presence of a higher molecular mass preprotein and a lower molecular mass band which probably corresponds to the mature protein exported to the periplasm. Immunoblotting performed with sera from either naturally infected or B. melitensis H38 experimentally infected sheep confirmed the importance of the B. melitensis CP28/BP26 protein as diagnostic antigen.     

DNA sequence analysis of X-ray-induced deletions at the white locus of Drosophila melanogaster

We have determined the nucleotide sequence of 5 X-ray-induced white mutants containing small rearrangements. Comparison with wild-type sequences showed deletions in the coding region ranging in size between 6 bp and 29 bp. These small deletions are distributed non-randomly over the white locus. Two mutants contain the same 29-bp deletion, while the other 3 deletions are clustered. All 5 deletions have occurred between 2 and 3 bp repeats. One of the repeats is preserved in the novel junction formed by the deletion. Our results suggest that recombinational processes may be involved in the generation of X-ray-induced deletions.     

Effects of DNA Heterologies on Bacteriophage λ Recombination

Previous studies of bacteriophage λ recombination have provided indirect evidence that substantial sequence nonhomologies, such as insertions and deletions, may be included in regions of heteroduplex DNA. However, the direct products of heterology-containing heteroduplex DNA--heterozygous progeny phage--have not been observed. We have constructed a series of small insertion and deletion mutations in the cI gene to examine the possibility that small heterologies might be accommodated in heterozygous progeny phage. Genetic crosses were carried out between λcI(-) Oam29 and λcI(+) Pam80 under replication-restricted conditions. Recombinant O(+)P(+) progeny were selected on mutL hosts and tested for cI heterozygosity. Heterozygous recombinants were readily observed with crosses involving insertions of 4 to 19 base pairs (bp) in the cI gene. Thus, nonhomologies of at least 19 bp can be accommodated in regions of heteroduplex DNA during λ recombination. In contrast, when a cI insertion or deletion mutation of 26 bp was present, few of the selected recombinants were heterozygous for cI. Results using a substitution mutation, involving a 26-bp deletion with a 22-bp insertion, suggest that the low recovery of cI heterozygotes containing heterologies of 26 bp or more is due to a failure to encapsidate DNA containing heterologies of 26 bp or more into viable phage particles.