Mobilization of vancomycin resistance by transposon-mediated fusion of a VanA plasmid with an Enterococcus faecium sex pheromone-response plasmid

A striking feature of recent outbreaks of vancomycin-resistant (Vm^R) enterococci is the apparent horizontal dissemination of resistance determinants. The plasmids pHKK702 and pHKK703 from Enterococcus faecium clinical isolate R7 have been implicated in the conjugal transfer of Vm^R. pHKK702 is a 41-kb plasmid that contains an element indistinguishable from the glycopeptide-resistance transposon Tn1546. pHKK703 is an approx. 55-kb putative sex pheromone-response plasmid that is required for conjugative mobilization of pHKK702. During experiments in which strain R7 was used as a donor, a highly conjugative Vm^R transconjugant was isolated that formed constitutive cellular aggregates. Restriction analyses and DNA hybridizations revealed that the transconjugant harbored a single plasmid of approx. 92 kb and this plasmid (pHKK701) was composed of DNA from both pHKK702 and pHKK703. Results from DNA sequence analyses showed that a 39-kb composite transposon (Tn5506) from pHKK702 had inserted into pHKK703. The left end of Tn5506 contained a single insertion sequence (IS) element, IS1216V2, whereas the right end was composed of a tandem IS structure consisting of the novel 1065-bp IS1252 nested within an IS1216V1 element. Transposition of Tn5506 from pHKK702 to pHKK703 created an 8-bp target sequence duplication at the site of insertion and interrupted an ORF (ORFX) that was 91% identical to that of prgX, a gene proposed to negatively regulate sex pheromone response of the E. faecalis plasmid, pCF10. We propose that the interruption of ORFX by Tn5506 led to the constitutive cellular aggregation phenotype and thereby enhanced the efficiency with which Vm^R was transferred. Similar IS1216V-mediated transposition events may contribute to the horizontal spread of glycopeptide resistance among enterococci in nature.     

Characterization of plasmids from plant pathogenic pseudomonads

Physical characterization of the resident plasmids from Pseudomonas tabaci, P. angulata, and P. coronafaciens strains indicated that they harbored five different plasmid DNA species. Two ATCC strains of P. tabaci contained indistinguishable plasmids that we have named pJP1 and pJP2. An isolate of one of these strains contained a spontaneous variant of pJP1, pJP1¹, which contains an insertion of 3.9 Mdal. This 3.9-Mdal region did not hybridize to pJP1 indicating that the region was foreign DNA and not a duplication of a segment of DNA already present in pJP1. Another P. tabaci strain, PT27881, contained a third plasmid species, pJP27, which had few similarities to pJP1 or pJP2, but was indistinguishable from the plasmids from all three P. angulata strains. pJP27 and pJP1 had a small region, 8.8 Mdal, of sequence homology. The one strain of P. coronafaciens examined contained a plasmid, pJP50, which was different from the P. tabaci plasmids, but had the 8.8-Mdal region and additional regions of sequence homology with pJP1 and pJP27 as well as homology with a portion of the pJP1¹ insertion. A fourth strain of P. tabaci, PTBR-2, a pathogen on beans, contained plasmid pBW, the only plasmid that lacked detectable regions of homology with the other plasmids.     

Site-specific Bacterial Chromosome Engineering: ΦC31 Integrase Mediated Cassette Exchange (IMCE)

The bacterial chromosome may be used to stably maintain foreign DNA in the mega-base range¹. Integration into the chromosome circumvents issues such as plasmid replication, plasmid stability, plasmid incompatibility, and plasmid copy number variance. This method uses the site-specific integrase from the Streptomyces phage (Φ) C31^2,3. The ΦC31 integrase catalyzes a direct recombination between two specific DNA sites: attB and attP (34 and 39 bp, respectively)⁴. This recombination is stable and does not revert⁵. A "landing pad" (LP) sequence consisting of a spectinomycin- resistance gene, aadA (SpR), and the E. coli ß-glucuronidase gene (uidA) flanked by attP sites has been integrated into the chromosomes of Sinorhizobium meliloti, Ochrobactrum anthropi, and Agrobacterium tumefaciens in an intergenic region, the ampC locus, and the tetA locus, respectively. S. meliloti is used in this protocol. Mobilizable donor vectors containing attB sites flanking a stuffer red fluorescent protein (rfp) gene and an antibiotic resistance gene have also been constructed. In this example the gentamicin resistant plasmid pJH110 is used. The rfp gene⁶ may be replaced with a desired construct using SphI and PstI. Alternatively a synthetic construct flanked by attB sites may be sub-cloned into a mobilizable vector such as pK19mob⁷. The expression of the ΦC31 integrase gene (cloned from pHS62⁸) is driven by the lac promoter, on a mobilizable broad host range plasmid pRK7813⁹.A tetraparental mating protocol is used to transfer the donor cassette into the LP strain thereby replacing the markers in the LP sequence with the donor cassette. These cells are trans-integrants. Trans-integrants are formed with a typical efficiency of 0.5%. Trans-integrants are typically found within the first 500-1,000 colonies screened by antibiotic sensitivity or blue-white screening using 5-bromo-4-chloro-3-indolyl-beta-D-glucuronic acid (X-gluc). This protocol contains the mating and selection procedures for creating and isolating trans-integrants.     

Construction and characterization of a plasmid containing a nearly full-size DNA copy of bacteriophage MS2 RNA.

An apparently full-length complementary DNA copy of in vitro polyadenylated MS2 RNA was synthesized with avian myeloblastosis virus RNA-dependent DNA polymerase. After the MS2 RNA template was removed from the complementary DNA strand with T₁ and pancreatic RNase digestion, the complementary DNA became a good template for the synthesis of double-stranded MS2 DNA with Escherichia coli DNA polymerase I. We then constructed molecular chimeras by inserting the double-stranded MS2 DNA into the PstI restriction endonuclease cleavage site of the E. coli plasmid pBR322 by means of the poly(dA)· poly(dT) tailing procedure. An E. coli transformant carrying a plasmid with a nearly full-length MS2 DNA insertion, called pMS2-7, was chosen for further study. Correlation between the restriction cleavage site map of pMS2-7 DNA and the cleavage map predicted from the primary structure of MS2 RNA, and nucleotide sequence analysis of the 5′ and 3′ end regions of the MS2 DNA insertion, showed that the entire MS2 RNA had been faithfully copied, and that, except for 14 nucleotides corresponding to the 5′-terminal sequence of MS2 RNA, the fulllength DNA copy of the viral genetic information had been inserted into the plasmid. Restriction endonuclease analysis of the chimera plasmid DNA also revealed the presence of an extra DNA insertion which was identified as the translocatable element IS1³ (see following paper).     

Plasmid pHH1 of Halobacterium salinarium: characterization of the replicon region,the gas vesicle gene cluster and insertion elements

The DNA sequence of the 5.7 kb plasmid pHH9 containing the replicon region of the 150 kb plasmid pHH1 from Halobacterium salinarium was determined. The minimal region necessary for stable plasmid maintenance lies within a 2.9 kb fragment, as defined by transformation experiments. The DNA sequence contained two open reading frames arranged in opposite orientations, separated by an unusually high AT-rich (60–70% A + T) sequence of 350 bp. All H. salinarium strains (H. halobium, H. cutirubrum) investigated harbour endogenous plasmids containing the pHH1 replicon; however, these pHH1-type plasmids differ by insertions and deletions. Adjacent to the replicon, and separated by a copy of each of the insertion elements ISH27 and ISH26, is the 9 kb p-vac region required for gas vesicle synthesis. Analysis of these and other ISH element copies in pHH1 revealed that most of them lack the target DNA duplication usually found with recently transposed ISH elements. These results underline the plasticity of plasmid pHH1.     

Isolation and identification of TL-DNA/plant junctions in Convolvulus arvensis transformed by Agrobacterium rhizogenes strain A4

We have constructed a Charon 4A phage library containing insert DNA isolated from a morning glory (Convolvulus arvensis) plant (clone 7) regenerated from a root organ culture incited by Agrobacterium rhizogenes, strain A4. Using a subcloned region of the Ri plasmid as ³²P-labeled probe, two lambda clones containing most of the 'left' T-DNA (TL) region were isolated. One of these lambda clones contains the left TL-DNA/plant junction, which was located by comparing nucleotide sequences from the appropriate regions of the Ri plasmid and this lambda clone. A 25-bp sequence found near this left TL-DNA/plant junction matches the 25-bp terminal sequence found at or near T-DNA/plant junctions of both nopaline- and octopine-type A. tumefaciens Ti plasmids. A possible location for the right Ri TL-DNA/plant junction in C. arvensis clone 7 was found by obtaining the nucleotide sequence surrounding its mapped location. Hybridization of plant DNA found adjacent to the left TL-DNA/plant junction against total C. arvensis DNA shows that this T-DNA integration occurred in a plant DNA region that does not contain highly repetitive DNA sequences. Nucleotide sequence analysis of 1004 bp of this plant DNA revealed no complete or partial open reading frames, but this plant DNA does have the potential to form various secondary structures which might play a role in the T-DNA integration event.     

DNA repair and sequence context affect 1O2-induced mutagenesis in bacteria

Electronic excited molecular oxygen (singlet oxygen, ¹O₂) is known to damage DNA, yielding mutations. In this work, the mutagenicity induced by ¹O₂ in a defined sequence of DNA was investigated after replication in Escherichia coli mutants deficient for nucleotide and base excision DNA repair pathways. For this purpose a plasmid containing a ¹O₂-damaged 14 base oligonucleotide was introduced into E.coli by transfection and mutations were screened by hybridization with an oligonucleotide with the original sequence. Mutagenesis was observed in all strains tested, but it was especially high in the BH20 (fpg), AYM57 (fpg mutY) and AYM84 (fpg mutY uvrC) strains. The frequency of mutants in the fpg mutY strain was higher than in the triple mutant fpg mutY uvrC, suggesting that activity of the UvrABC excinuclease can favor the mutagenesis of these lesions. Additionally, most of the mutations were G→T and G→C transversions, but this was dependent on the position of the guanine in the sequence and on repair deficiency in the host bacteria. Thus, the kind of repair and the mutagenesis associated with ¹O₂-induced DNA damage are linked to the context of the damaged sequence.     

A high security double lock and key mechanism in HUH relaxases controls oriT-processing for plasmid conjugation

Relaxases act as DNA selection sieves in conjugative plasmid transfer. Most plasmid relaxases belong to the HUH endonuclease family. TrwC, the relaxase of plasmid R388, is the prototype of the HUH relaxase family, which also includes TraI of plasmid F. In this article we demonstrate that TrwC processes its target nic-site by means of a highly secure double lock and key mechanism. It is controlled both by TrwC–DNA intermolecular interactions and by intramolecular DNA interactions between several nic nucleotides. The sequence specificity map of the interaction between TrwC and DNA was determined by systematic mutagenesis using degenerate oligonucleotide libraries. The specificity map reveals the minimal nic sequence requirements for R388-based conjugation. Some nic-site sequence variants were still able to form the U-turn shape at the nic-site necessary for TrwC processing, as observed by X-ray crystallography. Moreover, purified TrwC relaxase effectively cleaved ssDNA as well as dsDNA substrates containing these mutant sequences. Since TrwC is able to catalyze DNA integration in a nic-site-containing DNA molecule, characterization of nic-site functionally active sequence variants should improve the search quality of potential target sequences for relaxase-mediated integration in any target genome.     

Primer-directed mutagenesis of linearized plasmids

Two novel mutagenesis techniques to specifically alter the sequence of plasmid DNA have been developed. In contrast to other primer-directed mutagenesis methods which require a single-stranded, closed-circular template, a linearized single strand was used as the mutagenesis template. The template is prepared by restriction enzyme digestion of covalently-closed-circular plasmid DNA. These methods are simple, require small amounts of plasmid DNA, and can result in a relatively high frequency of mutagenesis.     

Novel mutations of Escherichia coli that produce recombinogenic lesions in DNA: V. Recombinogenic plasmids from arl mutants of Escherichia coli are unusually sensitive to nuclease S1 and partially deficient in cytosine methylation at C-C-(A/T)-G-G sequences

Two novel phenotypes previously associated with arl mutations of Escherichia coli, increased frequencies of genetic recombination and unusual sensitivity of DNA to the single-strand-specific nuclease S₁, have been defined most completely by the properties of λ bacteriophages grown on arl bacteria (Arl^? phages). We now find that plasmids maintained in arl mutants (Arl^? plasmids) exhibit elevated recombination frequencies, unusual sensitivity to nuclease S₁ (in a limited number of regions) and a new Arl phenotype, partially deficient methylation of the inner cytosine at C-C-(A/T)-G-G sequences.A variety of Arl^? plasmids (all pBR322 derivatives) show elevated recombination (4 to 10-fold) by three different assays (frequencies of homomultimers and of heteromultimers, efficiency of intramolecular recombination). Plasmids from arl bacteria (after conversion to linear form) are nicked by nuclease S₁ about 0.7 times per duplex; Arl⁺ plasmids are nuclease S₁-resistant. Restriction endonuclease EcoRII (recognition sequence, C-C-(A/T)-G-G) cuts Arl^? plasmid DNA more readily than Arl⁺ DNA, but Arl^? plasmids are still more EcoRII-resistant than Dcm^? plasmids (from E. coli dcm mutants, which lack the chromosomal cytosine methylase; recognition sequence, also C-C-(A/T)-G-G). By chromatographic analyses, Arl^? plasmid DNA contains less 5-methylcytosine than Arl⁺ (0.07% versus 0.15%). although the 6-methyladenine content is the same (0.5mol%).     

Ultra-Low Background DNA Cloning System

Yeast-based in vivo cloning is useful for cloning DNA fragments into plasmid vectors and is based on the ability of yeast to recombine the DNA fragments by homologous recombination. Although this method is efficient, it produces some by-products. We have developed an “ultra-low background DNA cloning system” on the basis of yeast-based in vivo cloning, by almost completely eliminating the generation of by-products and applying the method to commonly used Escherichia coli vectors, particularly those lacking yeast replication origins and carrying an ampicillin resistance gene (Amp^r). First, we constructed a conversion cassette containing the DNA sequences in the following order: an Amp^r 5′ UTR (untranslated region) and coding region, an autonomous replication sequence and a centromere sequence from yeast, a TRP1 yeast selectable marker, and an Amp^r 3′ UTR. This cassette allowed conversion of the Amp^r-containing vector into the yeast/E. coli shuttle vector through use of the Amp^r sequence by homologous recombination. Furthermore, simultaneous transformation of the desired DNA fragment into yeast allowed cloning of this DNA fragment into the same vector. We rescued the plasmid vectors from all yeast transformants, and by-products containing the E. coli replication origin disappeared. Next, the rescued vectors were transformed into E. coli and the by-products containing the yeast replication origin disappeared. Thus, our method used yeast- and E. coli-specific “origins of replication” to eliminate the generation of by-products. Finally, we successfully cloned the DNA fragment into the vector with almost 100% efficiency.     

Cloning of restriction and modification genes in E. coli: The HhaII system from Haemophilus haemolyticus

The genes for a Class II restriction-modification system (HhaII) from Haemophilus haemolyticus have been cloned in Escherichia coli. The vector used for cloning was plasmid pBR322 which confers resistance to tetracycline and ampicillin and contains a single endonuclease R·PstI site, (5′)C-T-G-C-A^↓-G (3′), in the ampicillin gene. The procedure developed by Bolivar et al. (1977) was used to form DNA recombinants. H. haemolyticus DNA was cleaved with PstI endonuclease and poly(dC) extensions were added to the 3′-OH termini using terminal deoxynucleotidyl transferase. Circular pBR322 DNA was cleaved to linear molecules with PstI endonuclease and poly(dG) extensions were added to the 3′-OH termini, thus regenating the PstI cleavage site sequence. Recombinant molecules, formed by annealing the two DNAs, were used to transfect a restriction and modification-deficient strain of E. coli (HB101 r^?m^?recA). Tetracycline-resistant clones were tested for acquisition of restriction phenotype (as measured by growth on plates seeded with phage λcI·O). A single phage-resistant clone was found. The recombinant plasmid, pDI10, isolated from this clone, had acquired 3 kilobases of additional DNA which could be excised with PstI endonuclease. In addition to the restriction function, cells carrying the plasmid expressed the HhaII modification function. Both activities have been partially purified by single-stranded DNA-agarose chromatography. The cloned HhaII restriction activity yields cleavage patterns identical to HinfI. A restriction map of the cloned DNA segment is presented.     

Excision of a DNA sequence determining kanamycin resistance from a ColE1-Km recombinant plasmid

Summary A 4.8×10⁶ dalton ECoRI-generated fragment of the R-factor R6-5 carrying the gene for kanamycin resistance (Km) was joined in vitro to ECoRI-treated ColE1 plasmid DNA. Transformation ofE. coli with the ColE1-Km recombinant plasmid yielded clones, which were immune to colicin E1, resistant to kanamycin and failed to produce colicin E1. During multiplication of this recombinant plasmid in the presence of chloramphenicol, cells expressed an increased resistance to kanamycin. Transformation studies with the recombinant DNA molecule showed very frequent loss of Km resistance in those cells harbouring a preexisting F'gal plasmid. Since colicin immunity is not affected and the col^- phenotype is still present, one has to test for a remaining DNA sequence further existing in ColE1 DNA by cleaving the plasmid DNA with the ECoRI restriction endonuclease. The full length of ColE1 DNA (6.2 kb) was restored, which confirmed that no deletion of ColE1 DNA sequences had occured. The remaining DNA sequence was identified as a 2.0 or 2.2 kb segment. On the basis of the length of the excised fragment it is proposed that the insertion sequence IS1 and a part of the inverted repeat sequence with coordinates 21.0 to 22.0 of the R6-5 DNA are recognised by a nucleolytic function.     

Identification of the translocatable element IS1 in a molecular chimera constructed with plasmid pBR322 DNA into which a bacteriophage MS2 DNA copy was inserted by the poly(dA).poly(dT) linker method.

Analysis of the plasmid DNA derived from a colony of bacteria carrying pMS2-7 (preceding paper, Devos et al., 1979) revealed the presence of an additional, smaller plasmid DNA, identified as pMS2-701. It was shown that pMS2-701 also contained the nearly full-length MS2 DNA copy, but the extra DNA insertion that was present to the right of the MS2 DNA insert in pMS2-7 was absent. Transformation of Escherichia coli with a DNA preparation containing both plasmid DNAs allowed the recloning of the pMS2-701 plasmid. Upon further subculturing, however, the pMS2-7 type plasmid containing the extra DNA insertion reappeared. Furthermore, the proportion of pMS2-7 relative to pMS2-701 increased in the course of successive subcultures. The extra DNA insertion in pMS2-7 was identified as the translocatable element IS1³ by mapping of restriction sites and by nucleotide sequence analysis. The boundaries between IS1 and pMS2-7 DNA revealed that IS1 had been inserted between the N-proximal part of the ampicillin gene and the poly(dA)-poly(dT) linker, and that a repetitious sequence of nine base-pairs in length had been generated by the translocation process.     

Selection and timing of replication of plasmids R1drd-19 and F'lac in Escherichia coli.

The selection and timing of plasmid replication was studied in exponentially growing cultures of Escherichia coli K-12 carrying the plasmid R1drd-19 and E. coli strains B/r A and B/r F carrying the plasmid F′lac. In all cases plasmid replication was studied by analysis of covalently closed circular (CCC) DNA. The turnover time of replicating plasmid DNA into CCC-DNA was found to be less than 4 min. Density shift experiments (from ¹⁵NH₄⁺, D₂O to ¹⁴NH₄⁺, H₂O) showed that plasmids R1drd-19 and F′lac are selected randomly for replication. This means that one of the plasmid copies in a cell is selected and replicated. There is no further plasmid replication in the cell until all plasmid copies, including the newly formed ones, have the same probability of being selected for replication. The early kinetics of the appearance of light plasmid DNA after the density shift showed that the time interval between successive replications of plasmids R1drd-19 and F′lac is $\text{τ}\text{n}$, where τ is the generation time and n is the average number of plasmid replications per cell and cell cycle. In a second type of experiment, exponentially growing cells were separated into a series of size classes by low-speed centrifugation in sucrose step gradients. Replication of plasmids R1drd-19 and F′lac was equally frequent in all size classes. This result is in accordance with the results of the density shift experiment. It can therefore be concluded that replication of plasmids R1drd-19 and F′lac is evenly spread over the whole cell cycle, which means that one plasmid replication occurs every time the cell volume has increased by one initiation mass.     

Development of a novel site-specific mutagenesis assay using MALDI-ToF MS (SSMA-MS)

We have developed and validated a novel site-specific mutagenesis assay, termed SSMA-MS, which incorporates MALDI-ToF mass spectrometry (MALDI-MS) analysis as a means of determining the mutations induced by a single DNA adduct. The assay involves ligating an adducted deoxyoligonucleotide into supF containing pSP189 plasmid. The plasmid is transfected into human Ad293 kidney cells allowing replication and therefore repair or a mutagenic event to occur. Escherichia coli indicator bacteria are transformed with recovered plasmid and plasmids containing the insert are identified colormetrically, as they behave as frameshift mutations. The plasmid is then amplified and digested using a restriction cocktail of Mbo11 and Mnl1 to yield 12 bp deoxyoligonucleotides, which are characterized by MALDI-MS. MALDI-MS takes advantage of the difference in molecular weight between bases to identify any induced mutations. This analysis method therefore provides qualitative and quantitative information regarding the type and frequency of mutations induced. This assay was developed and validated using an O⁶-methyl-2′-deoxyguanosine adduct, which induced the expected GC→AT substitutions, when replicated in human or bacterial cells. This approach can be applied to the study of any DNA adduct in any biologically relevant gene sequence (e.g. p53) in human cells and would be particularly amenable to high-throughput analysis.     

A comparison of virulence determinants in an octopine Ti plasmid, a nopaline Ti plasmid, and an Ri plasmid by complementation analysis of Agrobacterium tumefaciens mutants

Transposon-insertion mutants with vir^? Ti plasmids were characterized and then used in complementation experiments. One of the mutants (LBA 1517) had a mutation in a newly discovered vir locus called virF. The virF mutation led to a strongly diminished virulence on tomato and tobacco, but not on certain other plant species. Also a mutant (LBA 1505) was isolated with a mutation somewhere in the bacterial genome but outside the octopine Ti plasmid that caused a restriction in host range for tumor induction. Introduction of a nopaline Ti plasmid or an Ri plasmid into LBA 1505 did not restore normal virulence, showing that the vir gene affected in LBA 1505 determines a factor which is essential for normal tumor induction both by different types of Ti plasmids and by the Ri plasmid. The introduction of R primes containing part or all of the octopine Ti plasmid virulence region led to a restoration of virulence in strains with a vir^? nopaline Ti plasmid. Also the transfer of an Ri plasmid to a large number of different vir^? octopine or nopaline Ti plasmid mutants rendered these strains virulent. These results indicate that the octopine Ti plasmid, the nopaline Ti plasmid, and the Ri plasmid each have a similar virulence system which can mediate the transfer of T-DNA to plant cells from different types of Ti or Ri plasmids. In complementation experiments between vir^? octopine Ti plasmid mutations and vir^? nopaline Ti plasmid mutations it was found that equivalent functions are determined by the areas of DNA homology in the virulence regions of these two types of Ti plasmids. The previously defined octopine Ti plasmid virC locus appeared to consist of two different loci. One of these loci was found to be in a region of the octopine Ti plasmid which does not share DNA homology with the nopaline Ti plasmid, and was therefore called virO (octopine Ti plasmid specific). For the other locus the name virC was retained. Whereas mutations in the virC locus were avirulent on all plant species tested, mutations in virO were avirulent on tomato and pea, but virulent on sunflower and Nicotiana rustica. VirO^? mutants produced rooty tumors on Kalanchoë tubiflora.     

Analysis of N-acetoxy-N-2-acetylaminofluorene-induced frame-shifts in pBR322

We present a simple system which can be used to study directly directly the sequence change and the cellular repair functions involved in frame-shift mutagenesis by a covalently reactive mutagen. Positive (+S) and negative (?S) alterations in the number of base pairs of the Tc gene of pBR322 were generated and particular clones with Ap^RTc^S phenotypes were selected for mutagenesis experiments. Exposure of these frame-shifted plasmid DNAs to a potent carcinogen, N-acetoxy-N-2-acetylaminofluorene (AAAF), in vitro, caused covalent alterations to DNA sequence and resulted in a number of revertants (Ap^RTc^R) not observed in the untreated controls. The dose curve indicated an exponential response suggesting single-hit kinetics. Differential inactivation of the Ap gene was observed among various E. coli strains. The wild-type AB1157 and AB2463 (yrecA) showed a similar dose curve while AB1886 (uvrA) showed a marked decrease in Ap^R clones at the same dose. Both addition (+S) and deletion (?S) plasmids exhibited similar dose curves on inactivation of Ap gene. The reversion frequency, however, of ?S plasmid was a factor of 10 times higher than +S plasmid. The reversion frequency also increase markedly with uvrA host but not with recA host. 2 types of deletion revertants of the +S plasmid were found. 1 revertant has a single GC base-pair deletion in GC-rich region which is likely to be a target for AAAF reaction. The other showed a deletion of 4 base pairs (TCGA) at the tandem repeating sequence TCGATCGA which may represent a hot spot for frame-shift mutation.