The nucleotide sequence of IS5 from Escherichia coli

Investigating How Genomic Contexts Impact IS5 Transposition Within the Escherichia coli Genome

Insertions of the transposable element IS5 into its target sites in response to stressful environmental conditions, DNA structures, and DNA-binding proteins are well studied, but how the genomic contexts near IS5's native loci impact its transpositions is largely unknown. Here, by examining the roles of all 11 copies of IS5 within the genome of E. coli strain BW25113 in transposition, we reveal that the most significant copy of IS5 is one nested within and oriented in the same direction as the nmpC gene, while two other copies of IS5 harboring point mutations are hardly transposed. Transposition activity is heavily reliant on the upstream nmpC promoter that drives IS5 transposase gene ins5A, with more transpositions resulting from greater promoter activity. The IS5 element at nmpC but not at other loci transcribed detectable amounts of ins5A mRNA. By increasing expression of the ins5CB operon harbored in IS5, we demonstrate that Ins5B and Ins5C appear to exert a stimulatory role in IS5 transposition, suggesting that the downstream genomic regions near the native loci are involved in overall IS5 transposition as well. Using a strain that carries IS5 only at the nmpC locus, we confirm that IS5 primarily uses a copy/paste mechanism for transposition, although we cannot rule out the cut/paste mechanism.

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.

Visualizing evolution in real time to determine the molecular mechanisms of n-butanol tolerance in Escherichia coli

Toxicity of products or feedstock components poses a challenge in the biocatalyst-based production of fuels and chemicals. The genetic determinants that are involved in increased resistance to an inhibitor form the adaptive landscape for the phenotype; so in order to engineer more robust biocatalysts, a better understanding of the adaptive landscape is required. Here, we used an adaptive laboratory evolution method called visualizing evolution in real time (VERT) to help map out part of the adaptive landscape of Escherichia coli tolerance to the biofuel n-butanol. VERT enables identification of adaptive events (population expansions triggered by adaptive mutants) via visualization of the relative proportions of different fluorescently-labeled cells. Knowledge of the occurrence of adaptive events allows for a more systematic isolation of adaptive mutants while simultaneously reducing the number of missed adaptive mutants (and the underlying adaptive mechanisms) that result from clonal interference during the course of in vitro evolution. Based on the evolutionary dynamics observed, clonal interference was found to play a significant role in shaping the population structure of E. coli during exposure to n-butanol, and VERT helped to facilitate the isolation of adaptive mutants from the population. We further combined adaptive laboratory evolution with genome shuffling to significantly enhance the desired n-butanol tolerance phenotype. Subsequent transcriptome analysis of the isolated adaptive mutants revealed different mechanisms of n-butanol resistance in different lineages. In one fluorescently-marked subpopulation, members of the Fur regulon were upregulated; which was not observed in the other subpopulation. In addition, genome sequencing of several adaptive mutants revealed the genetic basis for some of the observed transcriptome profiles. We further elucidated the potential role of the iron-related gene in n-butanol tolerance via overexpression and deletion studies and hypothesized that the upregulation of the iron-related genes indirectly led to modifications in the outer membrane, which contributed to enhanced n-butanol tolerance.

Species-specific typing of DNA based on palindrome frequency patterns

DNA in its natural, double-stranded form may contain palindromes, sequences which read the same from either side because they are identical to their reverse complement on the sister strand. Short palindromes are underrepresented in all kinds of genomes. The frequency distribution of short palindromes exhibits more than twice the inter-species variance of non-palindromic sequences, which renders palindromes optimally suited for the typing of DNA. Here, we show that based on palindrome frequency, DNA sequences can be discriminated to the level of species of origin. By plotting the ratios of actual occurrence to expectancy, we generate palindrome frequency patterns that allow to cluster different sequences of the same genome and to assign plasmids, and in some cases even viruses to their respective host genomes. This finding will be of use in the growing field of metagenomics.

Designing and engineering evolutionary robust genetic circuits

Background

One problem with engineered genetic circuits in synthetic microbes is their stability over evolutionary time in the absence of selective pressure. Since design of a selective environment for maintaining function of a circuit will be unique to every circuit, general design principles are needed for engineering evolutionary robust circuits that permit the long-term study or applied use of synthetic circuits.

Results

We first measured the stability of two BioBrick-assembled genetic circuits propagated in Escherichia coli over multiple generations and the mutations that caused their loss-of-function. The first circuit, T9002, loses function in less than 20 generations and the mutation that repeatedly causes its loss-of-function is a deletion between two homologous transcriptional terminators. To measure the effect between transcriptional terminator homology levels and evolutionary stability, we re-engineered six versions of T9002 with a different transcriptional terminator at the end of the circuit. When there is no homology between terminators, the evolutionary half-life of this circuit is significantly improved over 2-fold and is independent of the expression level. Removing homology between terminators and decreasing expression level 4-fold increases the evolutionary half-life over 17-fold. The second circuit, I7101, loses function in less than 50 generations due to a deletion between repeated operator sequences in the promoter. This circuit was re-engineered with different promoters from a promoter library and using a kanamycin resistance gene (kanR) within the circuit to put a selective pressure on the promoter. The evolutionary stability dynamics and loss-of-function mutations in all these circuits are described. We also found that on average, evolutionary half-life exponentially decreases with increasing expression levels.

Conclusions

A wide variety of loss-of-function mutations are observed in BioBrick-assembled genetic circuits including point mutations, small insertions and deletions, large deletions, and insertion sequence (IS) element insertions that often occur in the scar sequence between parts. Promoter mutations are selected for more than any other biological part. Genetic circuits can be re-engineered to be more evolutionary robust with a few simple design principles: high expression of genetic circuits comes with the cost of low evolutionary stability, avoid repeated sequences, and the use of inducible promoters increases stability. Inclusion of an antibiotic resistance gene within the circuit does not ensure evolutionary stability.

Insertion sequence elements of Pseudomonas savastanoi: Nucleotide sequence and homology with Agrobacterium tumefaciens transfer DNA

Two types of transposable elements, IS51 and IS52 (IS, insertion sequence), were found in Pseudomonas syringae subsp. savastanoi (P. savastanoi) that spontaneously insert into and inactivate iaaM; the insertion results in the loss of indoleacetic acid production and attenuation of virulence. The nucleotide sequences of both IS elements have sizes and structural features common to other prokaryotic IS elements; IS51 is 1311 base pairs (bp) long and has terminal inverted repeats of 26 bp; IS52 is 1209 bp long and has terminal inverted repeats of 10 bp with a 1 bp mismatch. In the insertion involving IS51, the trinucleotide sequence CAG is duplicated within iaaM sequences at the recombination junction; in those involving IS52 the tetranucleotide sequences TTAG or CTAG are duplicated within iaaM sequences at the recombination junction. A copy of IS51 occurs 2.5 kilobases downstream from IaaH. In contrast to the high copy number of IS51 in the genome of the bacterium, only a few copies of IS52 are present. No nucleotide sequence homology was found between IS51 and IS52. However, a striking nucleotide sequence homology was found between a 531-bp region of IS51 and a portion of the central region of transfer DNA (T-DNA) in the octopine plasmid pTi15955 from Agrobacterium tumefaciens. These observations, together with our earlier finding on the homology between iaaM and iaaH and between gene 1 and gene 2 of transfer DNA, further suggest that genes for indoleacetic acid production in the two systems have a common origin.

Genome-wide localization of mobile elements: experimental, statistical and biological considerations

BACKGROUND

The distribution and location of insertion elements in a genome is an excellent tool to track the evolution of bacterial strains and a useful molecular marker to distinguish between closely related bacterial isolates. The information about the genomic locations of IS elements is available in public sequence databases. However, the locations of mobile elements may vary from strain to strain and within the population of an individual strain. Tools that allow de novo localization of IS elements and are independent of existing sequence information are essential to map insertion elements and advance our knowledge of the role that such elements play in gene regulation and genome plasticity in bacteria.

RESULTS

In this study, we present an efficient and reliable method for linear mapping of mobile elements using whole-genome DNA microarrays. In addition, we describe an algorithm for analysis of microarray data that can be applied to find DNA sequences physically juxtaposed with a target sequence of interest. This approach was used to map the locations of the IS5 elements in the genome of Escherichia coli K12. All IS5 elements present in the E. coli genome known from GenBank sequence data were identified. Furthermore, previously unknown insertion sites were predicted with high sensitivity and specificity. Two variants of E. coli K-12 MG1655 within a population of this strain were predicted by our analysis. The only significant difference between these two isolates was the presence of an IS5 element upstream of the main flagella regulator, flhDC. Additional experiments confirmed this prediction and showed that these isolates were phenotypically distinct. The effect of IS5 on the transcriptional activity of motility and chemotaxis genes in the genome of E. coli strain MG1655 was examined. Comparative analysis of expression profiles revealed that the presence of IS5 results in a mild enhancement of transcription of the flagellar genes that translates into a slight increase in motility.

CONCLUSION

In summary, this work presents a case study of an experimental and analytical application of DNA microarrays to map insertion elements in bacteria and gains an insight into biological processes that might otherwise be overlooked by relying solely on the available genome sequence data.

Transcript analysis ofEscherichia coliK-12 insertion element IS5

The mobile insertion element IS5 is a relatively small but genetically compact DNA sequence of 1195bp found in variable copy number in the genome of Escherichia coli strains. This study presents a detailed transcript analysis of the population of IS5 elements present in E. coli strains MC4100 and MG1655. The findings indicate that the ins5A gene comprising 978bp is transcribed from its own promoter, which is located close to the right-hand end of the element. The two divergently transcribed genes ins5C and in5B form an operon, and this transcript is fully contained within the borders of the ins5A transcript. Although transcription out of IS5 from element-internal promoters was negligible, in the case of MG1655 a major transcript was found to extend into the insertion element. This suggests that IS5-specific transcription can be influenced by the specific location of the element in the chromosome, the orientation it adopts and the gene it interrupts.

Functional analysis of the Escherichia coli molybdopterin cofactor biosynthesis protein MoeA by site-directed mutagenesis

Five moeA mutants were generated by replacing some conserved amino acids of MoeA by site-directed mutagenesis. The mutants were assayed for the ability to restore in vivo nitrate reductase activity of the moeA mutant Escherichia coli JRG97 and in vitro Neurospora crassa nit-1 nitrate reductase activity. The replacements Asp59AlaGly60Ala, Asp259Ala, Pro298AlaPro301Ala abolished the function of MoeA in Mo-molybdopterin formation and stabilization, reflected in the inability to restore nitrate reductase activity. The replacements Gly251AlaGly252Ala reduced, and that of Pro283Ala had no effect, on nitrate reductase activity. E. coli JRG97 cells transformed with mutants that failed to restore nitrate reductase activity showed by HPLC analysis a decreased level of molybdopterin-derived dephospho FormA as compared to bacteria transformed with wild-type moeA. The effects of the amino acid replacements on MoeA function may be explained in correlation with the MoeA crystal structure.

Sequence analysis of the family of penicillinase-producing plasmids of Neisseria gonorrhoeae

The exact nature of the sequence differences between the medically important family of gonococcal penicillinase-producing plasmids has been ascertained. The entire DNA sequence of the Asia-type plasmid, pJD4, demonstrated that it is 7426 bp and contains two direct repeats (DR30) that are implicated in the formation of deletion variant plasmids, such as the Africa-type plasmid. We have identified putative DnaA and IHF binding sites, various open reading frames that are thought to specify functional proteins, and some important DNA sequences involved with conjugative transfer of gonococcal beta-lactamase plasmids. The deletion in the Africa-type plasmid is 1827 bp and one of the DR30 repeats is also missing. The deletion in the Rio-type plasmid and several Toronto-type plasmids was determined to be 2273 bp and the sequence spanning the deletion was identical irrespective of geographic or temporal origin. The &Ncirc;imes-type plasmid is an Africa-type plasmid and also contains an IS5 insertion sequence. Since IS5 has not been identified in gonococcal isolates, we suggest that this sequence may have been inserted after the original gonococcal plasmid was transformed into Escherichia coli. The New Zealand plasmid is an Asia-type plasmid that contains an endogenous tandem duplication of 1883 bp and the direct DR2 is implicated in this duplication. The nature of the defined truncation of Tn2 present in the various plasmids is also discussed.

The tdcE gene in Escherichia coli strain W3110 is separated from the rest of the tdc operon by insertion of IS5 elements

Unlike other Escherichia coli K-12 strains, W3110 contains multiple copies of the insertion sequence IS5. Some of these IS5 elements have been involved in tandem duplication of a portion of the chromosome which includes, amonst others, the tdcABC-DEFG operon genes. The nucleotide sequence and insertion site of one of these elements, IS5P, was determined. It was shown that IS5P has inserted within the coding sequence of the tdcA gene and is flanked, not by the remaining portion of the tdcA gene, but by the extreme 3' end of the tdcD gene. In other E. coli K-12 strains the tdcD gene and three other genes, tdcE, tdcF and tdcG, all form part of the tdc operon. Our results demonstrate that during the duplication event the tdcABCgenes have been amplified and separated from the remaining genes tdcE, tdcF and tdcG of the operon, which are each present in single copy.

Analysis of genes involved in biosynthesis of coronafacic acid, the polyketide component of the phytotoxin coronatine

Coronafacic acid (CFA) is the polyketide component of coronatine (COR), a phytotoxin produced by the plant-pathogenic bacterium Pseudomonas syringae. The genes involved in CFA biosynthesis are encoded by a single transcript which encompasses 19 kb of the COR gene cluster. In the present study, the nucleotide sequence was determined for a 4-kb region located at the 3' end of the CFA biosynthetic gene cluster. Three open reading frames were identified and designated cfa8, cfa9, and tnp1; the predicted translation products of these genes showed relatedness to oxidoreductases, thioesterases, and transposases, respectively. The translational products of cfa8 and cfa9 were overproduced in Escherichia coli BL21; however, tnp1 was not translated in these experiments. Mutagenesis and complementation analysis indicated that cfa8 is required for the production of CFA and COR. Analysis of a cfa9 mutant indicated that this gene is dispensable for CFA and COR production but may increase the release of enzyme-bound products from the COR pathway; tnp1, however, had no obvious function in CFA or COR biosynthesis. A genetic strategy was used to produce CFA in a P. syringae strain which lacks the COR gene cluster; this approach will be useful in future studies designed to investigate biosynthetic products of the CFA gene cluster.

Novel HPV types present in oral papillomatous lesions from patients with HIV infection

Patients infected with the human immunodeficiency virus (HIV) often develop multiple papillomatous lesions of the oral cavity. In the present study, a total of 67 biopsies from benign oral lesions were analyzed for the presence of human papillomavirus (HPV) DNA using Southern-blot hybridization in combination with a polymerase chain reaction designed to detect all known HPV types, as well as unidentified types. These samples, collected at random from a high-risk population, were subsequently divided into 57 biopsies originating from patients with confirmed HIV infection and 10 biopsies from patients with unknown HIV status. Each sample was amplified with 7 different combinations of degenerate primers. All amplified products were sequenced. HPV DNA sequences were detected in 67% (45/67) of the samples. HPV 7 (19%) and HPV 32 (28%) were the predominant HPV types. HPV 32 was present in 2/4 fibromas tested. Two new HPV types, HPV 72 and HPV 73, were identified in oral warts with atypia. The complete genomes of these viruses were cloned and sequenced. Other HPV types detected were HPV 2a, HPV 6b, HPV 13, HPV 16, HPV 18, HPV 55, HPV 59 and HPV 69.

Genetics of alkane oxidation by Pseudomonas oleovorans

Many Pseudomonads are able to use linear alkanes as sole carbon and energy source. The genetics and enzymology of alkane metabolism have been investigated in depth for Pseudomonas oleovorans, which is able to oxidize C5-C12 n-alkanes by virtue of two gene regions, localized on the OCT-plasmid. The so-called alk-genes have been cloned in pLAFR1, and were subsequent analyzed using minicell expression experiments, DNA sequencing and deletion analysis. This has led to the identification and characterization of of the alkBFGHJKL and alkST genes which encode all proteins necessary to convert alkanes to the corresponding acyl-CoA derivatives. These then enter the beta-oxidation-cycle, and can be utilized as carbon- and energy sources. Medium (C6-C12)- or long-chain (C13-C20) n-alkanes can be utilized by many strains, some of which have been partially characterized. The alkane-oxidizing enzymes used by some of these strains (e.g. two P. aeruginosa strains, a P. denitrificans strain and a marine Pseudomonas sp.) appear to be closely related to those encoded by the OCT-plasmid.

Isolation and characterization of the Rickettsia prowazekii recA gene

The recA gene has been isolated from Rickettsia prowazekii, an obligate intracellular bacterium. Comparison of the amino acid sequence of R. prowazekii RecA with that of Escherichia coli RecA revealed that 62% of the residues were identical. The highest identity was found with RecA of Legionella pneumophila, in which 69% of the residues were identical. Amino acid residues of E. coli RecA associated with functional activities are conserved in rickettsial RecA, and the R. prowazekii recA gene complements E. coli recA mutants for UV light and methyl methanesulfonate sensitivities as well as recombinational deficiencies. The characterized region upstream of rickettsial recA did not contain a sequence homologous to an E. coli LexA binding site (SOS box), suggesting differences in the regulation of the R. prowazekii recA gene.

Regulation of the catechol 1,2-dioxygenase- and phenol monooxygenase-encoding pheBA operon in Pseudomonas putida PaW85

In Pseudomonas putida PaW85, the ortho-cleavage pathway is used for catechol degradation. The 11.4-kb XhoI fragment cloned from phenol degradation plasmid pEST1226 into pKT240 (recombinant plasmid pAT1140) contains the inducible pheBA operon that encodes catechol 1,2-dioxygenase (gene pheB) and phenol monooxygenase (gene pheA), the first two enzymes for the phenol degradation pathway. The promoter of the pheBA operon is mapped 1.5 kb upstream of the pheB gene. The plasmid pAT1140, when introduced into P. putida PaW85, enables the bacteria to use the hybrid plasmid-chromosome-encoded pathway for phenol degradation. The synthesis of the plasmid-encoded phenol monooxygenase and catechol 1,2-dioxygenase is induced by cis,cis-muconate. The expression studies of the deletion subclones derived from pAT1140 revealed that the transcription of the pheBA operon is positively controlled by a regulatory protein that is chromosomally encoded in P. putida. cis,cis-Muconate in cooperation with positive transcription factor CatR activates the transcription of the chromosomal ortho-pathway genes catA and catBC in P. putida (R. K. Rothmel, T. L. Aldrich, J. E. Houghton, W. M. Coco, L. N. Ornston, and A. M. Chakrabarty, J. Bacteriol. 172:922-931, 1990). The inability to express the pheBA operon in a P. putida CatR- background and activation of transcription of the pheBA operon in Escherichia coli in the presence of the catR-expressing plasmid demonstrated that the transcription of the pheBA operon in P. putida PaW85 carrying pEST1226 is controlled by the chromosomally encoded CatR.

Characterization of the spontaneous elimination of streptomycin sensitivity (SmS) on high-copy-number plasmids: SmS-enforcement cloning vectors with a synthetic rpsL gene

The strAS or rpsL+ gene, encoding a ribosomal protein, S12, expresses its streptomycin-sensitivity (SmS) phenotype dominantly over strAR or rpsL- gene. Therefore, strAR cells that harbor plasmids with strAS alleles are phenotypically SmS. It was found that the SmS phenotype is unstable, and such cells eventually switch to the Sm-resistance (SmR) phenotype, especially when the strAS gene was cloned on high-copy-number (HCN) plasmids. It seemed that the strA gene cloned on HCN plasmids was toxic to Escherichia coli host cells and, during prolonged cultivation, plasmids with an inactivated strAS gene, mostly carrying insertion sequence elements, such as IS1, IS5 and gamma delta, were selected. The instability of the strA gene was particularly enhanced when the Val51 residue in the middle of S12 protein was replaced by Leu, suggesting enhanced toxicity of the altered S12. Since the strAS gene was stably maintained throughout approx. 100 cell doublings when its expression was abolished, most probably it is the gene product rather than the nucleotide sequence itself that is responsible for the instability of strA gene on HCN plasmids. To improve the stability of the SmS phenotype, the previously reported ampicillin-resistance-conferring and SmS-enforcing plasmid vector, pHSG670, was reconstructed. The resulting vector, pHSG683, confers chloramphenicol resistance, enforces SmS on strAR and supE- host bacteria, and has multiple cloning sites within the coding region of synthetic rpsL gene. When pHSG683 DNA was prepared from strAR and sup+ cells grown in tryptophan-rich medium with Cm and Sm, less than 10(-6) plasmids failed to enforce SmS on strAR and supE- cells in tryptophan-less medium with Cm.(ABSTRACT TRUNCATED AT 250 WORDS)

IS5: a mobile enhancer of transcription in Escherichia coli

The cryptic bgl operon of Escherichia coli is activated by the spontaneous insertion of mobile DNA elements. Screening of a collection of such mutations revealed insertion of the 1195-base-pair element IS5 into various positions both upstream and downstream of the bgl promoter P0. Activation of the operon was in all cases attributable to enhancement of P0 activity. Introduction of internal deletions into IS5 almost completely abolished P0 enhancement, demonstrating that enhancement is not simply the result of mutational inactivation of some inhibitory sequences. Intact copies of IS5 in trans restored the enhancing activity of the deletion derivatives. The trans-activator is encoded by IS5 gene ins5A, an essential transposition function. Activation of gene expression by means of interaction of a defective mobile element in cis with functions encoded by a nondefective element in trans has so far been described only for a maize controlling element.

Mapping of insertion element IS5 in the Escherichia coli K-12 chromosome. Chromosomal rearrangements mediated by IS5

We identified phage clones containing insertion element IS5 in a set of 476 lambda phage clones carrying chromosomal segments that cover almost the entire chromosome of Escherichia coli K-12 W3110. Precise locations and orientations of IS5 were then determined by cleavage analysis of phage DNAs containing them. We mapped 23 copies of IS5 (named is5A to is5W) on the W3110 chromosome. Among them, ten were identified as the common elements present at the same locations in both chromosomes of W3110 and another E. coli K-12 strain, JE5519. While most of the mapped IS5 elements were scattered over the W3110 chromosome, four copies of IS5 (designated is5L, is5M, is5N and is5O) were in a region representing tandem duplication of a DNA segment flanked by two copies of IS5. Interestingly, one unit of this DNA segment as well as a portion of it was seen also in a tandem array in a different region where two copies of IS5 (designated is5P and is5Q) were present. In particular two pairs of the mapped IS5 elements may have been involved in inversion of the chromosomal segments in two of the E. coli K-12 derivatives.

The complete AvrII restriction map of the Escherichia coli genome and comparisons of several laboratory strains

The complete 13 site AvrII restriction map of the genome of E coli strain MG1655 is presented and compared with several other E. coli strains. The map was determined primarily by isolating individual AvrII fragments from pulsed-field gels, and hybridizing these large probes to a battery of mapped E. coli clones in lambda vectors. AvrII restriction patterns for eight other laboratory strains were determined and maps for seven of them deduced from the gel and comparisons between the strain genotypes, the MG1655 map, and AvrII sites in E. coli sequences taken from Genbank.

Genetic analysis of the yopE region of Yersinia spp.: identification of a novel conserved locus, yerA, regulating yopE expression

The yopE gene of Yersinia pseudotuberculosis was recently sequenced, and YopE was identified as an indispensable virulence determinant when tested in a mouse model (A. Forsberg and H. Wolf-Watz, Mol. Microbiol. 2:121-133, 1988). In the study described here, the DNA sequences of the yopE genes of Yersinia pestis EV76 and Yersinia enterocolitica 8081 were determined and compared with that of the Y. pseudotuberculosis gene. Only two codons were found to differ, both leading to amino acid replacements, when the gene from Y. pestis was compared. These two replacements were also present in the gene from Y. enterocolitica; in addition, 18 other codons were found to differ. Thirteen of these substitutions led to amino acid replacements. Downstream of the yopE gene, the plasmid partition locus par was found to be conserved in all three species. In Y. enterocolitica 8081, the sequence homology was interrupted by a putative insertion sequence element inserted between the yopE gene and the par region at a position only 5 base pairs downstream of the yopE stop codon. Upstream of the yopE gene, 620 base pairs were conserved in the three species. This region contained a 130-amino-acid-long open reading frame reading in the opposite direction to the yopE gene and expressed a 14-kilodalton protein in minicells. An insertion mutation in this region constructed in Y. pseudotuberculosis expressed significantly lower amounts of YopE protein in vitro than did the corresponding wild type. The expression level could be restored by transcomplementation. This new locus was designated yerA, for yopE-regulating gene A. The yerA mutant was avirulent when mice were challenged by oral infection.

Gene amplification contributes to sulfonamide resistance in Escherichia coli

A sulfathiazole-resistant strain of Escherichia coli was isolated and shown to contain a fourfold tandemly amplified segment of DNA 18 kilobase pairs in length in addition to a mutationally altered dihydropteroate synthase, the target enzyme for sulfonamide inhibition. The amplified DNA contained a gene designated sur that contributed to sulfathiazole resistance when present in greater amounts than those in the wild type. Sulfathiazole resistance was markedly decreased upon loss of the amplified DNA after nonselective growth. Plasmids that contained sur also conferred only weak sulfathiazole resistance on wild-type strains. Comparison of the restriction maps of the amplified DNA, wild-type DNA, and sur-containing plasmids showed that a DNA rearrangement occurred before or concomitant with the DNA amplification event. The DNA rearrangement resulted from an IS5 insertion, which, in conjunction with an IS5 element residing near sur in the wild-type strain, resulted in an -IS5-sur-IS5- configuration. Homologous recombination could account for duplication and subsequent amplification of the sur region. High-copy-number plasmids containing the sur locus did not express a sulfathiazole-resistant dihydropteroate synthase, nor did they overexpress wild-type dihydropteroate synthase. These data suggest that the high level of sulfathiazole resistance in this strain results from a synergistic effect of two different mutations.

Constitutive activation of the fucAO operon and silencing of the divergently transcribed fucPIK operon by an IS5 element in Escherichia coli mutants selected for growth on L-1,2-propanediol

L-1,2-Propanediol is an irretrievable end product of L-fucose fermentation by Escherichia coli. Selection for increased aerobic growth rate on propanediol results in the escalation of basal synthesis of the NAD+-linked oxidoreductase encoded by fucO, a member of the fuc regulon for the utilization of L-fucose. In general, when fucO becomes constitutively expressed, two other simultaneous changes occur: the fucA gene encoding fuculose-1-phosphate aldolase becomes constitutively expressed and the fucPIK operon encoding fucose permease, fucose isomerase, and fuculose kinase becomes noninducible. In the present study, we show that fucO and fucA form an operon which is divergently transcribed from the adjacent fucPIK operon. In propanediol-positive and fucose-negative mutants the cis-controlling region shared by the operons fucAO and fucPIK is lengthened by 1.2 kilobases. DNA hybridization identified the insertion element to be IS5. This element, always oriented in the same direction with the left end (the BglII end) proximal to fucA, apparently causes constitutive expression of fucAO and noninducibility of fucPIK. The DNA of the fucAO operon and a part of the adjacent fucP was sequenced.

Streptokinase mutations relieving Escherichia coli K-12 (prlA4) of detriments caused by the wild-type skc gene

A novel phenotype is described for Escherichia coli K-12 carrying the prlA4 allele determining a membrane component of the protein export mechanism. It is manifest as transformation deficiency for plasmids containing the cloned group C streptococcal streptokinase gene, skc. Streptokinase plasmid mutations relieving the prlA4 strain of this deficiency fell into three classes. Class 1 included skc::IS5 insertions, with IS5 integrated in a region encoding the Skc signal sequence and inactivating skc. Class 2 included IS1 insertions leaving skc intact but reducing skc expression, presumably by altering the function of the skc promoter as judged by an insertion site close to the -35 region. The most interesting class, 3, included skc deletions removing the entire signal sequence or a tetrapeptide from its hydrophobic core. The tetrapeptide deletion reduced the size, hydrophobicity, and predicted alpha-helicity of the central region of the Skc signal sequence but facilitated the export of mature Skc in both the wild type and the prlA4 mutant. These findings indicate that the incompatibility between prlA4 and skc is related to deleterious effects of the Skc signal sequence. The tetrapeptide deletion may function by altering the conformation of the signal sequence so as to render interaction with both the PrlA wild-type protein and the PrlA4 mutant protein less detrimental to the export mechanism. These findings also provide an explanation for the difficulties encountered in cloning streptokinase genes in E. coli plasmids and maintaining their structural stability.

Use of synthetic ribosome binding site for overproduction of the 5B protein of insertion sequence IS5

Insertion sequence IS5 is a bacterial transposable element which contains three open reading frames designated 5A, 5B and 5C. Although there was no detectable expression from the 5B open reading frame when it was preceded by the native promoter and ribosome binding site or by a tac promoter and the native ribosome binding site, we have overproduced a 5B protein both in vitro and in Escherichia coli cells by using a tac promoter and a specially-designed synthetic ribosome binding site. beta-galactosidase fusion studies suggested that the synthetic binding site is at least 150-fold more efficient than the native binding site. The 5B protein amounted to 80-85% of the total protein made in vitro and 20-25% of the total protein pulse-labelled in whole cells. It is stable in vitro but rapidly degraded in vivo. Thus expression of the 5B gene appears to be limited by both poor translation initiation and protein degradation.

Overproduction and purification of the M.HhaII methyltransferase from Haemophilus haemolyticus

The HhaII methyltransferase gene from Haemophilus haemolyticus was subcloned in an expression vector under control of the hybrid trp-lac promoter. Induction with isopropyl-beta-D-thiogalactopyranoside results in overproduction of the methyltransferase to about 3% of total cellular protein. The methyltransferase was purified to near electrophoretic homogeneity by phosphocellulose, DEAE, and gel chromatography. Its monomer Mr by sodium dodecyl sulfate-polyacrylamide gel electrophoresis is 25 kDa, in good agreement with that predicted from the nucleotide sequence. Crystals of the methyltransferase were obtained in the presence of a two-fold molar excess of the duplex oligodeoxynucleotide substrate 5'd-GGACTCC.CCTGAGG.

A selection cartridge for rapid detection and analysis of spontaneous mutations including insertions of transposable elements in Enterobacteriaceae

We present a method that allows positive selection and rapid analysis of mutations in Enterobacteriaceae. Mutations are detected in a 2630 bp selection cartridge inserted in two different bacterial multicopy plasmid vectors. Spontaneous mutations in Escherichia coli, Enterobacter cloacae and Citrobacter freundii include insertions, deletions and point mutations. The small size of the target sequence facilitates rapid analysis of DNA rearrangements by cleavage with restriction enzymes and of any type of mutation by DNA sequence analysis. While in E. coli insertions of the mobile elements IS1, IS2 and IS5 were readily found, insertions of putative new transposable elements were detected in Enterobacter cloacae. The selection cartridge can thus serve as a tool for studying the spectrum of insertion mutations in Enterobacteriaceae and probably other Gram-negative bacteria, and the dependency of this spectrum on physiological and environmental factors and the host's genetic background can be investigated.

Synthesis and overproduction of the 5A protein of insertion sequence IS5

We have demonstrated both the synthesis and overproduction of the 5A protein encoded by the longest open reading frame of the bacterial insertion sequence IS5. Expression was obtained in vitro and in Escherichia coli maxicells from plasmids containing IS5 in either orientation, as well as in vitro from a restriction fragment containing exclusively IS5 DNA. When IS5 was cloned in the appropriate orientation downstream of a strong tac promoter, production of the 5A protein was increased to 10 to 20% of the total protein synthesized in vitro.

Insertion sequence IS5 contains a sharply curved DNA structure at its terminus

It was demonstrated that insertion sequence IS5 contains a sequence-directed bent (sharply curved) DNA structure at its terminus, close to one of its 16 bp terminal repeats. The minimal number of copies of IS5 related sequences and the locations of the latter on the Escherichia coli K12 W3110 chromosome were determined. Evidence is presented of the occurrence of IS5 mediated translocation and duplication of a large DNA segment on the E. coli chromosome.

Second-element turn-on of gene expression in an IS1 insertion mutant

To learn more about the ways in which genes silenced by insertion mutations can be reactivated, we have undertaken a systematic investigation of Gal+ revertants of the polar mutant galOP-306::IS1 in Escherichia coli K12. The selective conditions used excluded reversion to wild type by precise excision of IS1. In this system (which resided on a multi-copy plasmid) reversion to the Gal+ phenotype occurred with a frequency of about 10(-7) per cell and per generation. Analysis of the revertants revealed that - with the single exception of the previously published chromosomal mutant sis1 - alterations in the structure of IS1 lead to reactivation of gal operon expression. These events fall into four classes: (I) insertion of IS2 at position 327 in IS1, insertion of IS2 at position 687 in IS1, (III) insertion of a hitherto undetected mobile element, IS150, at position 387, (IV) a 16-bp deletion encompassing IS1 coordinates 553-568. Of some 200 independent reversion events studied, all but one were of types I-III i.e. they involved the intervention of a second mobile element.

Growth-rate-dependent expression and cloning of gnd alleles from natural isolates of Escherichia coli

6-Phosphogluconate dehydrogenase (6PGD), encoded by gnd, is highly polymorphic among isolates of Escherichia coli form natural populations. As a means of characterizing the growth-rate-dependent regulation of the level of 6PGD, five gnd alleles, including the E. coli B/r allele, were crossed into E. coli K-12 with bacteriophage P1. In each of the isogenic strains, the level of 6PGD was two- to threefold higher in cells grown on glucose than in cells grown on acetate. The level of enzyme activity in the acetate-grown cells varied about sixfold within the set of isogenic strains. The physiological importance of these differences in enzyme level is discussed. The gnd gene was cloned from five E. coli strains and Salmonella typhimurium LT-2 and mapped with twelve restriction endonucleases. gnd was located and oriented on the chromosomal DNAs. The restriction maps of the genes were aligned at conserved restriction sites, and the relative divergence of the genes was estimated from restriction site polymorphisms. The E. coli gnd genes differed from the S. typhimurium gene by about 11%. Most of the E. coli genes differed from one another by less than 5%, but one allele differed from the others by about 10%. Only the gnd gene from E. coli K-12 had an IS5 element located nearby.

Molecular characterization of the oligopeptide permease of Salmonella typhimurium

The oligopeptide permease (Opp) of Salmonella typhimurium is a periplasmic binding protein-dependent transport system and handles any peptides containing from two to five amino acid residues. Opp plays an important nutritional role and is also required for the recycling of cell wall peptides. We have determined the nucleotide sequence of the opp operon. In addition to the four opp genes identified previously by genetic means (oppABCD) a fifth gene, oppF, is shown to be cotranscribed as part of the opp operon. Using reverse genetics, we show that oppF also encodes an essential component of the Opp transport system. The five proteins, OppABCDF, are shown to be the only proteins required for Opp function. Regulation of opp expression and of the differential expression of genes within the operon is investigated. We have devised a simple means of constructing lacZ gene fusions to any S. typhimurium chromosomal gene in vivo, using derivatives of bacteriophage Mu. Using this procedure, opp-lacZ gene fusions were selected. The resultant Opp-LacZ hybrid proteins were used to show that OppB, OppC and OppD are membrane-associated proteins. A detailed comparison of the Opp components with those of other binding protein-dependent transport systems provides insight into the mechanisms and evolution of these transport systems.

Peptide transport in Salmonella typhimurium: molecular cloning and characterization of the oligopeptide permease genes

The oligopeptide permease is encoded by at least four genes which are transcribed as a single operon. We cloned and characterized this operon from Salmonella typhimurium, as well as the flanking genes, tonB, ana and a new gene, cwd, which affects cell wall synthesis. We correlated the physical map of opp DNA with a detailed genetic map of the opp operon and the individual opp genes were accurately located with respect to various restriction sites by Southern blotting. The region of the chromosome near opp was found to be highly unstable with deletions arising at a highly frequency. The operon also contains hot-spots for IS1 and IS5 insertions.

Identification and sequence analysis of a silent gene (ushA0) in Salmonella typhimurium

The evolution of new or improved enzyme specificities in prokaryotes has been proposed to involve gene duplication, followed by silencing of one of the duplicates at the transcriptional or translational level. Such "silent gene intermediates" are distinct from "cryptic" genes, which are proposed to have a different role in evolution. We describe the identification in Salmonella typhimurium of a silent gene (ushA0) using the active (homologous) ushA gene (encoding UDP-sugar hydrolase) from Escherichia coli as a probe. The ushA0 gene has been cloned and, in the multicopy state, very weak expression can be detected; the gene product was shown to be immunologically and functionally related to the enzyme from E. coli. The sequence of the ushA0 gene was found to be highly homologous to the previously determined sequence of the ushA gene, and the respective promoter and ribosomal-binding sites are also very similar. However, a presumed strong rho-independent terminator in the ushA gene is absent from ushA0; although a weak stem-and-loop structure is present in the 3' region of ushA0, its structure is atypical of rho-independent terminators. The sequence analysis also revealed an insertion-sequence like sequence at the 3' end of ushA0 with a convergent open reading frame terminating 116 base-pairs from the ushA0 stop codon. A deletion of the 5' region of the open reading frame results in increased expression of ushA0, indicating that convergent transcription plays some role in the silencing of ushA0. S. typhimurium contains a UDP-sugar hydrolase, biochemically and genetically distinct from that in E. coli, encoded by the ushB gene. Our results indicate that ushB is not strongly sequence-related to ushA0, and its gene product is not immunologically related to the ushA gene product. ushB is hence a functional duplicate of ushA and provides a rationale for the silencing of ushA0. This situation, and the DNA sequence comparison of ushA and ushA0, strongly suggests that rather than being a cryptic gene, ushA0 has been silenced recently during the evolution of S. typhimurium.

A naturally occurring large chromosomal inversion in Escherichia coli K12

Strain 1485IN and its derivatives were found to have a large inversion extending to about 35% of the chromosome. Because of this, the question arose as to whether 1485IN had arisen from an Escherichia coli strain other than K12. However, 1485IN had a flagellar antigen and a restriction-modification system indistinguishable from those of W3110, a major line of K12, and had retained an amber suppressor and lambda sensitivity that are characteristics of W1485 from which this strain seems to have arisen. Strain 1485IN had acquired proline auxotrophy, but showed the same growth rate as W1485 in nutrient broth at 37 degrees C. Interrupted matings with Hfr strains of 1485IN revealed a gene arrangement of nalA-gal-trp-his-lac-proA-thrleu-ilv, in which gal, trp, and his were on the inverted segment. The termini of the inversion were inferred to be situated between tsx (9.5 min) and purE (12 min) and between his (44 min) and cdd (46.5 min).

IS222, a new insertion element associated with the genome of Pseudomonas aeruginosa

A new insertion element, IS222, was identified to be associated with the DNA of a mutant strain of the converting Pseudomonas aeruginosa bacteriophage D3. The insertion sequence was 1,350 base pairs in size and possessed terminal inverted repeats. The nucleotide sequence contained single cleavage sites for EcoRI and PvuI but none for BamHI, PstI, HindIII, SmaI, or SalI. By Southern hybridization analysis, no homology was found with genomic DNA from P. aeruginosa PAT or Escherichia coli. Genomic DNA from the phage host, P. aeruginosa PAO, contained two sequences homologous to IS222.

Structural and functional studies of insertion element IS200

The nucleotide sequence of the insertion element IS200 has been determined partially, including the junctions between the element and the host chromosome at the insertion site. At most, two bases (A-A) are found repeated at the junctions and could be duplications of host sequences generated by the insertion of the element. No obvious sequence repeats, either direct or inverted, have been detected between the sequences just within the two ends of the element. The element is an extremely strong block to host transcription across the insertion site. A sequence similar to known transcription termination signals was found just within the element near the right end. Removal of less than 50 base-pairs at the right end of the element abolishes the transcription block. The putative terminator sequence is located within this 50 base-pair region. Genetic studies suggest that the element contains a promoter located more than 93 base-pairs from its left end. The proposed promoter and terminator are in proper orientation to form an internal transcription unit.

Improving the efficiency of dot-matrix similarity searches through use of an oligomer table

Dot-matrix sequence similarity searches can be greatly speeded up through use of a table listing all locations of short oligomers in one of the sequences to find potential similarities with a second sequence. The algorithm described finds similarities between two sequences of lengths M and N, comparing L residues at a time, with an efficiency of L X M X N/(SK) where S is the alphabet size, and k is the length of the oligomer. For nucleic acids, in which S = 4, use of a tetranucleotide table results in an efficiency of L X M X N/256. The simplicity of the approach allows for a straightforward calculation of the level of similarities expected to be found for given search parameters. Furthermore, the storage required is minimal, allowing for even large sequences to be compared on small microcomputers. Theoretical considerations regarding the use of this search are discussed.

Personal access to sequence databases on personal computers

A comprehensive package of software has been developed to access nucleic acid and protein sequence databases on stand-alone IBM personal computers. The software combines keyword search on the annotation fields of the data with pattern matching algorithms on the biological sequences. Sequences containing complex sites like promoters or kink sites can be identified as well as sequences that are similar to a query sequence. Protein sequences with particular patterns of amino acids such as hydrophobic regions can be identified as well. Considering the relatively inexpensive hard disks now available, personal computers have become a cost-effective alternative to mainframe processing for sequence databases.

ISL1: a new transposable element in Lactobacillus casei

The genome structures of a temperate Lactobacillus phage, phi FSW, and its virulent mutants, phi FSVs, were examined by restriction, heteroduplex and nucleotide-sequence analyses. The results showed that two out of three phi FSVs had the same 1.3 kbp insertion (designated as ISL1) at different positions in the phi FSW sequence. ISL1 was 1,256 bp long and contained at least two long open reading frames of 279 and 822 bases on one strand. Inverted repeats were found at the termini of the ISL1 which was bracketed by 3 bp direct repeats of the phi FSW sequence. From this evidence, we concluded that ISL1 was a transposable element in Lactobacillus casei.

Analysis of regulatory sequences upstream of the E. coli uvrB gene; involvement of the DnaA protein

A region located upstream of the uvrB promoters P1 and P2 was found to cause high plasmid loss when cloned in multicopy vectors. Two sequence elements responsible for this phenomenon were identified by mapping of spontaneous mutations that restore plasmid maintenance: a sequence known to have in vitro promoter activity and a partially overlapping sequence that shows extensive homology to recognition sites for the DnaA protein. Accordingly alterations in the level of DnaA protein in vivo were found to affect the extent of plasmid loss. A possible role for interaction of the DnaA protein with the region of interest is discussed in relation to regulation of uvrB expression.

Activation of expression of a cloned archaebacterial gene in Escherichia coli by IS2, IS5, or deletions

A DNA fragment from the methanogenic archaebacterium Methanococcus voltae, when cloned into the PstI site of the plasmid vector pBR322, complements the Escherichia coli argG mutation strongly or weakly depending on its orientation. Faster-growing variants derived from a strain containing the poorly expressed fragment were found to harbor plasmids which had undergone genetic rearrangements. Some of the plasmids were shown to have acquired an insertion element (IS2 or IS5), derived from the E. coli chromosome, close to the region essential for complementing activity. Other plasmids exhibited no homology with E. coli chromosomal DNA. These were found to represent multimeric forms of the parental plasmid in which 2-3 kb of DNA between the tet promoter and the argG-complementing region had been deleted. Growth rates of the variant strains in the absence of arginine varied significantly, suggesting differences in efficiency of activation of the cloned DNA.

IS30, a new insertion sequence of Escherichia coli K12

Three independent spontaneous mutations of prophage P1 affecting the ability of the phage to reproduce vegetatively are due to the insertion of a mobile genetic element, called IS30. The same sequence is also carried in the R plasmid NR 1-Basel, but not in the parental plasmid NR 1. Southern hybridisation study indicates that the Escherichia coli K 12 chromosome carries several copies of IS30 as a normal resident. IS30 is 1.2 kb long and contains unique restriction cleavage sites for BglII, ClaI, HindIII, NciI and HincII, and it is cleaved twice by the enzymes HpaII and TaqI. The ends of IS30 are formed by 26 bp long inverted repeats with 3 bases mismatched. Upon transposition IS30 generates a duplication of only 2 bp of the target. The following observations suggest a pronounced specificity in target selection by IS30. In transposition to the phage P1 genome a single integration site was used three times independently, and in both orientations. A short region of sequence homology has been identified between the P1 and NR 1-Basel insertion sites. IS30 has mediated cointegration as well as deletion. The entire IS30 sequences were duplicated in the cointegrates between a pBR322 derivative containing IS30 and the genome of phage P1-15, and several loci on the P1-15 genome served as fusion sites, some of which were used more than once.

Distribution of DNA insertion element IS5 in natural isolates of Escherichia coli

DNA from Escherichia coli strains in a reference collection of 72 recent natural isolates (ECOR strains) and 25 natural isolates from the "pre-antibiotic" period 1930-1940 (Murray strains) were studied to determine the genomic abundance of insertion element IS5 and the size of genomic restriction fragments carrying sequences homologous to IS5. Among the ECOR strains, nearly two-thirds lack DNA sequences that hybridize with IS5, and one-half of the remainder have only one copy. Among strains in which IS5 is present, extensive variation in the size of IS5-bearing restriction fragments occurs, in many cases allowing distinction among strains that are judged to be nearly identical in genotype because of the identical electrophoretic mobility of the enzyme coded by each of 11 chromosomal loci. Among the Murray strains in which IS5 is present, the average number of elements per strain is larger, but not markedly so, than among recent isolates. Comparison between duplicate strains in the Murray collection suggests that the rate of accumulation of IS5 elements in prolonged storage in stab tubes corresponds to an apparent probability of transposition of approximately 0.008 +/- 0.002 per IS5 element per year. Because of the extensive genetic variation among strains, insertion elements such as IS5 would seem to be convenient genetic markers with which to detect recent common ancestry among strains.

Insertions of transposable elements in the promoter proximal region of the gene cluster for Escherichia coli H+-ATPase: 8 base pair repeat generated by insertion of IS1

A plasmid pKY159 (Yamaguchi and Yamaguchi 1983) carrying a promoter proximal portion of the gene cluster of the proton-translocating ATPase (H+-ATPase) of Escherichia coli causes growth inhibition of wild-type cells. Insertion of a transposable element in this plasmid released this inhibitory effect. In analyzing this inhibitory effect, we determined the insertion points at the nucleotide-sequence level of transposable elements on 30 independent derivatives of pKY159 . Insertions of IS1, IS5, and gamma delta were found between the promoter and the gene for a possible component of 14,000 daltons of the H+-ATPase. Of 31 insertions, 26 were of IS1 and were located at the same site, indicating that this site is a hotspot for IS1 insertion and that IS1 insertion is much more frequent than that of IS5 or gamma delta in this region. Four different sites for IS1 insertion were found; in two of these an 8 base pair (bp) duplicate of the target sequence ( AAAAACGT and AAACGTTG ) was generated, while in the other two a 9 bp duplicate was found. In all cases in this study the nucleotide sequence of IS1 was the same as that of IS1-K. In the two cases with an 8 bp duplicate in different sites, a common 6 bp sequence ( AAACGT ) was found. These results suggested that generation of the 8 bp duplicate is related to the common sequence rather than a mutation in IS1 suggested by Iida et al. (1981) and also suggested that the essential length of the duplicate is 8 bp or less than 8 bp. A 6 bp sequence ( GTGATG ) homologous to the end portion of IS1 was found at the hotspot , but not at other sites, suggesting that this homology contributed to the high frequency of IS1 insertion.(ABSTRACT TRUNCATED AT 250 WORDS)

Characterization of the new insertion sequence IS91 from an alpha-hemolysin plasmid of Escherichia coli

IS91 is a 1.85 kb insertion sequence originally resident in the alpha-hemolytic plasmid pSU233. The element was transposed sequentially from this plasmid to pACYC184, to R388, and to pBR322. Both cointegrates and simple insertions of the element were obtained. A detailed restriction enzyme map of the element is presented. This does not bear any relationship to the maps of previously described insertion sequences. Furthermore, hybridization between these sequences and IS91 could not be demonstrated. Deletion derivatives of IS91 were constructed which are unable to transpose. However, their transposition can be complemented in trans by wild-type elements. One of these deletion derivatives has been genetically labeled with a kanamycin resistance marker from Tn5. When this new element was complemented for transposition, only about 2% of the transposition products were cointegrates. Thus, the behavior of IS91 is better explained by transposition models that allow direct transposition.