Inverted repeats of Tn5 are transposable elements

Deep learning for HGT insertion sites recognition

Background

Horizontal Gene Transfer (HGT) refers to the sharing of genetic materials between distant species that are not in a parent-offspring relationship. The HGT insertion sites are important to understand the HGT mechanisms. Recent studies in main agents of HGT, such as transposon and plasmid, demonstrate that insertion sites usually hold specific sequence features. This motivates us to find a method to infer HGT insertion sites according to sequence features.

Results

In this paper, we propose a deep residual network, DeepHGT, to recognize HGT insertion sites. To train DeepHGT, we extracted about 1.55 million sequence segments as training instances from 262 metagenomic samples, where the ratio between positive instances and negative instances is about 1:1. These segments are randomly partitioned into three subsets: 80% of them as the training set, 10% as the validation set, and the remaining 10% as the test set. The training loss of DeepHGT is 0.4163 and the validation loss is 0.423. On the test set, DeepHGT has achieved the area under curve (AUC) value of 0.8782. Furthermore, in order to further evaluate the generalization of DeepHGT, we constructed an independent test set containing 689,312 sequence segments from another 147 gut metagenomic samples. DeepHGT has achieved the AUC value of 0.8428, which approaches the previous test AUC value. As a comparison, the gradient boosting classifier model implemented in PyFeat achieve an AUC value of 0.694 and 0.686 on the above two test sets, respectively. Furthermore, DeepHGT could learn discriminant sequence features; for example, DeepHGT has learned a sequence pattern of palindromic subsequences as a significantly (P-value=0.0182) local feature. Hence, DeepHGT is a reliable model to recognize the HGT insertion site.

Conclusion

DeepHGT is the first deep learning model that can accurately recognize HGT insertion sites on genomes according to the sequence pattern.

A MATLAB-based tool for accurate detection of perfect overlapping and nested inverted repeats in DNA sequences

SUMMARY

Palindromic sequences, or inverted repeats (IRs), in DNA sequences involve important biological processes such as DNA-protein binding, DNA replication and DNA transposition. Development of bioinformatics tools that are capable of accurately detecting perfect IRs can enable genome-wide studies of IR patterns in both prokaryotes and eukaryotes. Different from conventional string-comparison approaches, we propose a novel algorithm that uses a cumulative score system based on a prime number representation of nucleotide bases. We then implemented this algorithm as a MATLAB-based program for perfect IR detection. In comparison with other existing tools, our program demonstrates a high accuracy in detecting nested and overlapping IRs.

AVAILABILITY AND IMPLEMENTATION

The source code is freely available on (http://bioinfolab.miamioh.edu/bioinfolab/palindrome.php)

CONTACT

liangc@miamioh.edu or karroje@miamioh.edu

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

A novel Escherichia coli-derived mutation detected with the Big Blue cII mutant selectable assay

Transgenic mouse mutation detection systems allow investigation of the origins and mechanisms of mutation associated with exogenous and endogenous mutagen exposures in individual tissues and cell types. In the past, selection assays for transgenic mutants have been contaminated with nonmurine-derived mutations and assay validation is critical to ensure murine in vivo origins of mutations. This is critical in studies of spontaneous mutations and extrapolation to endogenous mammalian genes. Herein, we provide one measure of the contribution of Escherichia coli (E. coli)-derived mutations to the Big Blue(R) cII transgene mutant selection assay. We report the first direct evidence of an E. coli-derived cII mutation identified among mutations recovered in the cII selective assay. An E. coli transposable 5 (Tn5) element IS50R inverted repeat (1,534 bp) was identified at base pair 414 in the cII transgene and the insertion generated a 9 bp target site duplication typical of this type of transposition. The bacterial transposition occurred only once in the assay of 25 x 10(6) plaque forming units and sequencing of 1,177 cII mutants. The observed frequency of this type of mutation is 4 x 10(-8) in retrieved lambda phage and 8.5 x 10(-4) in harvested cII mutants and thus a very rare occurrence in typical analyses of spontaneous in vivo mutations. Given that the frequency of transposition is equal to, or an order of magnitude higher, than the frequency of point mutations in E. coli, this article provides excellent validation for the murine origins of mutations detected using the cII mutant selection assay.

Target site selection by the mariner-like element, Mos1

The eukaryotic transposon Mos1 is a class-II transposable element that moves using a "cut-and-paste" mechanism in which the transposase is the only protein factor required. The formation of the excision complex is well documented, but the integration step has so far received less investigation. Like all mariner-like elements, Mos1 was thought to integrate into a TA dinucleotide without displaying any other target selection preferences. We set out to synthesize what is currently known about Mos1 insertion sites, and to define the characteristics of Mos1 insertion sequences in vitro and in vivo. Statistical analysis can be used to identify the TA dinucleotides that are non-randomly targeted for transposon integration. In vitro, no specific feature determining target choice other than the requirement for a TA dinucleotide has been identified. In vivo, data were obtained from two previously reported integration hotspots: the bacterial cat gene and the Caenorhabditis elegans rDNA locus. Analysis of these insertion sites revealed a preference for TA dinucleotides that are included in TATA or TA x TA motifs, or located within AT-rich regions. Analysis of the physical properties of sequences obtained in vitro and in vivo do not help to explain Mos1 integration preferences, suggesting that other characteristics must be involved in Mos1 target choice.

Accessory DNAs in the bacterial gene pool: playground for coevolution

Chemostat studies of bacteria that harbour the prokaryotic transposable elements Tn5 and Tn10 and the temperate phages lambda, Mu, P1 and P2 have shown that these accessory DNA elements confer a selective advantage on their hosts. We propose that similar selective effects provided the initial impetus for the evolution of nascent accessory DNA elements in primitive bacterial populations. In subsequent evolution the elements acquired or perfected the 'selfish' characteristics of over-replication and horizontal transmission. Such selfish traits led to the dissemination of accessory DNAs among commensal strains, species and genera, genetically interconnecting them to create a 'commonwealth' of species that potentially share a common gene pool. The involvement of accessory DNAs in genetic exchange provides selection at the population level for refinement and diversification of the elements and for regulation of their replication, transposition and transfer among cells. The diversity of intracellular environments encountered by the elements imposes constraints on their evolution while at the same time altering the selection pressures operating on conventional chromosomal genes. This process of coevolution of accessory DNAs with the genomes of their diverse hosts has led to a unique population structure and mechanism of genetic exchange among bacteria, which constitutes the most effective adaptive strategy yet devised by selection.

A bifunctional DNA binding region in Tn5 transposase

Tn5 transposition is a complicated process that requires the formation of a highly ordered protein-DNA structure, a synaptic complex, to catalyse the movement of a sequence of DNA (transposon) into a target DNA. Much is known about the structure of the synaptic complex and the positioning of protein-DNA contacts, although many protein-DNA contacts remain largely unstudied. In particular, there is little evidence for the positioning of donor DNA and target DNA. In this communication, we describe the isolation and analysis of mutant transposases that have, for the first time, provided genetic and biochemical evidence for the stage-specific positioning of both donor and target DNAs within the synaptic complex. Furthermore, we have provided evidence that some of the amino acids that contact donor DNA also contact target DNA, and therefore suggest that these amino acids help define a bifunctional DNA binding region responsible for these two transposase-DNA binding events.

Molecular and kinetic comparison of the novel extended-spectrum beta-lactamases CTX-M-25 and CTX-M-26

CTX-M-25 is a novel extended-spectrum beta-lactamase isolated from a single Canadian Escherichia coli isolate. Susceptibility testing demonstrated that this enzyme confers resistance to both cefotaxime and ceftazidime, but the level of resistance was reduced with the addition of beta-lactamase inhibitors. The bla(CTX-M-25) gene was detected on a 111-kb plasmid. It is a member of the CTX-M-8 group and has the closest amino acid identity (99%; three amino acid substitutions) with CTX-M-26. The bla(CTX-M-26) gene was detected on a 100-kb plasmid isolated from a Klebsiella pneumoniae strain from the United Kingdom, and plasmid profiling revealed that it showed some homology to the bla(CTX-M-25)-harboring plasmid. Both CTX-M genes were located downstream of ISEcp1, although the copy upstream of bla(CTX-M-25) was disrupted by IS50-A. Comparative kinetic studies of recombinant CTX-M-25 and CTX-M-26 enzymes showed that CTX-M-25 has a higher level of ceftazidime hydrolysis (kcat values, 33 and 0.005 s(-1) for CTX-M-25 and CTX-M-26, respectively).

oriT-directed cloning of defined large regions from bacterial genomes: identification of the Sinorhizobium meliloti pExo megaplasmid replicator region

We have developed a procedure to directly clone large fragments from the genome of the soil bacterium Sinorhizobium meliloti. Specific regions to be cloned are first flanked by parallel copies of an origin of transfer (oriT) together with a plasmid replication origin capable of replicating large clones in Escherichia coli but not in the target organism. Supplying transfer genes in trans specifically transfers the oriT-flanked region, and in this process, site-specific recombination at the oriT sites results in a plasmid carrying the flanked region of interest that can replicate in E. coli from the inserted origin of replication (in this case, the F origin carried on a BAC cloning vector). We have used this procedure with the oriT of the plasmid RK2 to clone contiguous fragments of 50, 60, 115, 140, 240, and 200 kb from the S. meliloti pExo megaplasmid. Analysis of the 60-kb fragment allowed us to identify a 9-kb region capable of autonomous replication in the bacterium Agrobacterium tumefaciens. The nucleotide sequence of this fragment revealed a replicator region including homologs of the repA, repB, and repC genes from other Rhizobiaceae, which encode proteins involved in replication and segregation of plasmids in many organisms.

Simple and efficient generation in vitro of nested deletions and inversions: Tn5 intramolecular transposition

We have exploited the intramolecular transposition preference of the Tn 5 in vitro transposition system to test its effectiveness as a tool for generation of nested families of deletions and inversions. A synthetic transposon was constructed containing an ori, an ampicillin resistance (Ampr) gene, a multi-cloning site (MCS) and two hyperactive end sequences. The donor DNA that adjoins the transposon contains a kanamycin resistance (Kanr) gene. Any Amprreplicating plasmid that has undergone a transposition event (Kans) will be targeted primarily to any insert in the MCS. Two different size targets were tested in the in vitro system. Synthetic transposon plasmids containing either target were incubated in the presence of purified transposase (Tnp) protein and transformed. Transposition frequencies (Ampr/Kans) for both targets were found to be 30-50%, of which >95% occur within the target sequence, in an apparently random manner. By a conservative estimate 10(5) or more deletions/inversions within a given segment of DNA can be expected from a single one-step 20 microl transposition reaction. These nested deletions can be used for structure-function analysis of proteins and for sequence analysis. The inversions provide nested sequencing templates of the opposite strand from the deletions.

DNA binding and phasing analyses of Tn5 transposase and a monomeric variant

Both full-length Tn 5 transposase and a COOH-terminal truncated monomeric form of the protein,n369, have been shown to specifically bind end sequences at comparable affinities. In addition, both proteins distort the target sequence in a similar manner, as determined by a circular permutation assay. In this study,nEK54, a derivative ofn369 with a single amino acid substitution that significantly enhances binding activity, is used in further binding and bending studies along with full-length transposase. Phasing analysis has shown that distortion of the end sequences upon binding of full-length transposase and nEK54 protein is due in part to a protein-induced bend oriented towards the major groove. Because the center of transposase-induced bending maps to the extreme leftward end of the 19 bp consensus sequence, we examined the possibility that optimal protein binding requires additional upstream nucleotide contacts. Experiments presented here show that 9-10 nucleotides are needed upstream of +1 of the 19 bp sequence for efficient binding and this requirement can be met by either single-stranded or double-stranded DNA.

High frequency of independent IS50 transposition during Tn5 mutagenesis of Bordetella pertussis virulence-associated genes

Transposon Tn5-generated mutants of Bordetella pertussis were selected on the basis of their inability to bind the dye Congo red (Crb-). No mutants which were solely Crb- were found. Ten mutants were phenotypically equivalent to previously described strains with mutations in the virulence regulatory (bvg) locus and failed to express a range of virulence-associated factors. Two of these mutants were shown to have Tn5 insertions within the bvg locus, while another two mutants showed deletions in this regulatory region. Complementation studies indicated that the other six mutants had spontaneous mutations in the bvg locus, but with Tn5 inserted elsewhere in the chromosome. Several of the mutants, besides having a single Tn5 insertion, also showed additional IS50 insertions, indicating that the IS50 element contained within Tn5 had transposed independently. Such additional insertion events, which themselves would have the potential to cause mutation, could complicate the interpretation of mutant phenotypes which could thus arise from the insertional inactivation of more than one gene.

Escherichia coli umuDC mutants: DNA sequence alterations and UmuD cleavage

The products of the chromosomally encoded umuDC genes are directly required for mutagenesis in Escherichia coli. Strains with either umuD or umuC mutations are rendered phenotypically non-mutable. To ascertain the molecular basis of this non-mutability, we determined the DNA sequence alterations of seven chromosomal umuDC mutants. Six mutants (umuD1, umuD44, umuD77, umuC36, umuC25, and umuC104) were found to be single base-pair substitutions that resulted in missense mutations. The Tn5 transposon insertion mutation (umuC122) resulted in a missense mutation followed immediately by a termination codon, producing a truncated UmuC protein lacking 102 carboxyl-terminal amino acids. All of the mutations were found to reside in regions of the UmuD and UmuC proteins that share high homology with analogous proteins. Chemiluminescent immunoassays revealed that the umuD1, umuD44, and umuD77 mutations all resulted in a non-cleavable UmuD protein. Because UmuD cleavage is a prerequisite for mutagenesis, the lack of UmuD processing appears to be the molecular basis for the non-mutable phenotype in these strains. These studies re-emphasize the critical nature of the RecA-mediated cleavage of UmuD for inducible mutagenesis and provide insights into the functional domains of the UmuC protein.

Effects of variation of inverted-repeat sequences on reactions mediated by the transposase of Tn21

The frequencies of one-ended transposition and normal transposition of derivatives of Tn21 that contain mutant inverted-repeat sequences (IRs) have been measured. In general, there was a linear relationship between the log of the frequency of one-ended transposition of a mutant IR and the log of the frequency of normal transposition of an element flanked by a wild-type IR at one end and by the mutant IR at the other. This implied that one-ended and normal transposition share the rate-limiting step that determines the frequency of transposition and that both IRs are involved in the rate-limiting step in normal transposition. Surprisingly, it was found that only the outer 18 base pairs of the IR of Tn21 engaged accurately in both one-ended and normal transposition, at about 1% of the frequency of the wild-type IR.

Using transposon Tn5 insertions to sequence bacteriophage T4 gene 11

A simple and rapid method of creating an overlapping set of deletions in cloned DNA in preparation for sequencing has been developed. The method is based on a positive selection for Tn5 transposition into the cloned DNA fragment on a high-copy-number filamid, resolution of potential filamid dimers by filamentous phage infection, and the use of Tn5 both as a "portable" restriction enzyme site for in vitro DNA deletion and a "portable" sequencing primer binding site to initiate DNA sequencing reactions using a custom primer complementary to the outside ends of IS50. This new method has been utilized to sequence bacteriophage T4 gene 11, encoding the T4 baseplate protein gp11. The coding sequence of gene 11 is 657 bp in length, and predicts a primary structure of 219 amino acids that agrees well with the biochemical data previously obtained. DNA sequence around gene 11 suggests that the expression of genes 10, 11, and 12 of phage T4 are translationally coupled. Plasmids carrying deletions generated using this method have been used to map genetically five amber alleles of gene 11. These amber alleles were sequenced to confirm the proposed reading frame. The five amber alleles actually represent two different mutational changes at either codon 206 or 207, changing these adjacent glutamine codons to amber. The position of these amber alleles lends support to earlier studies identifying the carboxyl terminus of gp11 as essential in the interaction of P11 with baseplate protein P10 (Plishker and Berget, 1984).

Tn5-mediated transposition of plasmid DNA after transduction to Myxococcus xanthus

After coliphage P1-mediated transfer of Tn5-containing plasmid DNA from Escherichia coli to Myxococcus xanthus, transductants were identified which contained plasmid sequences integrated at many sites on the bacterial chromosome. The unaltered plasmid DNA sequences in these transductants were apparently flanked by intact Tn5 or IS50 sequences. These results suggest that Tn5-mediated transposition has occurred and provide a method for integrating plasmid DNA into the M. xanthus chromosome without the requirement for homologous recombination.

Isolation and characterization of pathogenicity genes of Pseudomonas syringae pv. tabaci

Pseudomonas syringae pv. tabaci BR2 produces tabtoxin and causes wildfire disease on tobacco and bean plants. Approximately 2,700 Tn5 insertion mutants of a plasmid-free strain, PTBR 2.024, were generated by using suicide plasmid pGS9. Of these Tn5 mutants, 8 were no longer pathogenic on tobacco plants and 10 showed reduced symptoms. All of the eight nonpathogenic mutants caused typical wildfire disease symptoms on bean plants. Two of the nonpathogenic mutants failed to produce tabtoxin. The eight nonpathogenic mutants have Tn5 insertions into different EcoRI and SalI restriction fragments. The EcoRI fragments containing Tn5 from the eight nonpathogenic mutants were cloned into vector pTZ18R or pLAFR3. A genomic library of the parent strain was constructed in the broad-host-range cosmid pLAFR3. Three different cosmid clones that hybridized to the cloned Tn5-containing fragment from one of the nonpathogenic mutants, PTBR 4.000, were isolated from the genomic library. These clones contained six contiguous EcoRI fragments (a total of 57 kilobases [kb]). A 7.2-kb EcoRI fragment common to all three restored pathogenicity to mutant PTBR 4.000. None of the six EcoRI fragments hybridized to Tn5-containing fragments from the other seven mutants. The 7.2-kb fragment was conserved in P. syringae pv. tabaci and P. syringae pv. angulata, but not in other pathovars or strains. Because the mutants retained pathogenicity on bean plants and because of the conservation of the 7.2-kb EcoRI fragment only in pathovars of tobacco, we suggest that genes on the fragment might be related to host specificity.

Mutational analysis of insertion sequence 50 (IS50) and transposon 5 (Tn5) ends

Insertion sequence 50 (IS50) transposition utilizes a 19-base-pair "outside" end and a 19-base-pair "inside" end in inverted orientation relative to each other, whereas transposon 5 (Tn5) transposition utilizes two inverted outside ends. The frequency of transposition events that involve an inside end is regulated 1000-fold by the host dam methylase system. The end sequence requirements for transposition and its regulation by dam methylase were analyzed in Escherichia coli by generating random single base pair mutations in either an IS50 inside end or outside end placed in inverted orientation with respect to an unmutagenized outside end. The mutations were then isolated, assayed for transposition phenotype, and sequenced. Mutations were isolated at 15 of the 19 sites in the outside end. All of these mutations except those at position 4 decreased transposition. Mutations at position 4 (which is the only nonidentical base pair in a region of homology between the outside and inside ends) had no effect on transposition. Mutations were isolated at 11 of the 19 sites in the inside end. All of these mutations, including one at position 4, decreased transposition in dam- cells. Mutations at position 10 (within a dam recognition sequence) and 2 (not within a dam recognition sequence) reduced the magnitude of dam regulation. A mutation within a dam recognition sequence adjacent to the required 19 base pairs of the inside end did not reduce the magnitude of dam regulation.

Temporal control of transposition in Tn5

IS50R is an insertion sequence associated with the transposon Tn5. IS50R carries the structural genes for two proteins; one (P1) is the Tn5 transposase, and the other (P2) is an inhibitor of transposition. These two proteins are translated from two different transcripts, m1 and m2. When bacteriophage lambda::IS50R DNA was introduced into a bacterial cell, m1 and m2 were initially at relative levels of about 1 to 2. As time progressed the amount of m1 fell, whereas the amount of m2 continued to increase, until after about 3 h the ratio of m1 to m2 was about 1 to 80. The temporal changes in the levels of these transcripts correlated with temporal changes in P1 and P2 levels and Tn5 transposition that have been documented in other studies. We measured the stability of the messages and showed that the differences in the levels of m1 and m2 must reflect real differences in the strengths of their promoters and that the changes in transcription kinetics are mediated by the dam methylation system of the cell and are not determined by IS50R products. Our results show that the 5' end of m2 is about twice as stable as that of m1, which raises the possibility that differential message stability does, in part, influence the ratio of inhibitor to transposase.

Transposition of IS91 does not generate a target duplication

We determined the DNA sequences surrounding the junctions of IS91 in two insertion derivatives: pSU234 (pACYC184::IS91) and pSU240 (pBR322::IS91). The termini of IS91 consist of two imperfect inverted repeats eight base pairs long. Their sequence is 5'-TCGAGTAGG. . . CCTATCGA-3'. Insertion of IS91 did not generate direct repetitions in the target DNAs.

The evolution of DNA sequences in Escherichia coli

It is proposed that certain families of transposable elements originally evolved in plasmids and functioned in forming replicon fusions to aid in the horizontal transmission of non-conjugational plasmids. This hypothesis is supported by the finding that the transposable elements Tn3 and gamma delta are found almost exclusively in plasmids, and also by the distribution of the unrelated insertion sequences IS4 and IS5 among a reference collection of 67 natural isolates of Escherichia coli. Each insertion sequence was found to be present in only about one-third of the strains. Among the ten strains found to contain both insertion sequences, the number of copies of the elements was negatively correlated. With respect to IS5, approximately half of the strains containing a chromosomal copy of the insertion element also contained copies within the plasmid complement of the strain.

Vector insertion mutagenesis of Rhizobium sp. strain ORS571: direct cloning of mutagenized DNA sequences

When the limited-host-range plasmid pVP2021 carrying Tn5 was mobilized into Rhizobium sp. strain ORS571 and stable acquisition of Tn5 was selected, ORS571 plasmid-genome cointegrates were exclusively obtained; direct Tn5 transposition was never observed. In every case, genomic cointegrates exhibited an additional (third) IS50 element that bordered VP2021 DNA sequences but maintained a single Tn5 element. Genomic cointegrates containing IS50 triplications were stable; neither phenotypic reversion nor resolution was detectable. Auxotrophic mutant strains (vector insertion mutants) were identified at expected frequencies among derivatives carrying ostensibly random genomic pVP2021 insertions; N2 fixation (Nif)-defective vector insertion mutants were observed among these derivatives at a frequency of 10(-3). The presence of integrated pVP2021 in ORS571 nif::VP2021 mutant genomes enabled VP2021 to constitute an endogenous cloning vector. After EcoRI or KpnI digestions, genomic nif::pVP2021 DNA sequences contiguous with integrated pVP2021 were directly cloned as new replicons without addition of an exogenous vector. Recombinant plasmids derived from two such nif::pVP2021 mutants hybridized to previously analyzed ORS571 Nif DNA sequences. Recombinant plasmid DNA and ORS571 Nif region DNA were found to be colinear; pVP2021 insertions could be accurately mapped. pVP2021 insertion-mutagenesis thus allows the direct cloning of ORS571 gene sequences for which mutant phenotypes can be selected or screened.

recA-independent recombination between repeated IS50 elements is not caused by an IS50-encoded function

Certain pBR322-related plasmids containing direct repeats of the insertion element IS50 appear to be unstable in recA Escherichia coli because smaller recombinant derivatives accumulate rapidly in plasmid DNA populations. We show here that (i) this instability is plasmid specific, but not IS50 specific; (ii) it is due to a detrimental effect exerted by these plasmids on bacterial growth; and (iii) the growth impairment is alleviated in cells harboring the smaller recombinant plasmids. Although a recent report had concluded that accumulation of recombinants reflected an IS50-specific recombination function, when correction is made for the relative growth rates of cells containing the parental and recombinant plasmids the evidence for such a recombination function disappears.

Spontaneous deletion of citrate-utilizing ability promoted by insertion sequences

The citrate utilization (Cit+) transposon Tn3411 was shown to be flanked by directly repeated sequences (IS3411L and IS3411R) by restriction enzyme analysis and electron microscope observation. Cit- deletion mutants were frequently found to be generated in pBR322::Tn3411 by intramolecular recombination between the two copies of IS3411. The flanking IS3411 elements of Tn3411 were shown to be functional insertion sequences by Tn3411-mediated direct and inverse transposition. Tn3411-mediated inverse transposition from pBR322::Tn3411 to the F-plasmid derivative pED100 occurred more efficiently than that of direct transposition of the Cit+ determinant. This was thought to be due to the differential transposability of IS3411L and IS3411R in the transposition process. The frequency of transposition of IS3411 marked with a chloramphenicol resistance determinant was much higher than IS3411-mediated cointegrate formation, suggesting that replicon fusions are not essential intermediates in the transposition process of Tn3411 or IS3411. Spontaneous deletions occurred with high frequency in recA hosts. The spontaneous deletion promoted by homologous recombination between two IS3411 elements in Tn3411 was examined with deletion mutants.

Sequences essential for transposition at the termini of IS50

The DNA sequences found repeated in opposite orientation at the ends of insertion (IS) elements are thought to contain sites at which transposase proteins act during transposition. Many elements have repeats of at least 15 base pairs (bp). Those of IS50 are quite short, however: just 8 of the first 9 bp. Functional tests had indicated that one end of IS50 is more effective in transposition than the other end and suggested that at least one of the recognition sites of IS50 extends beyond the common 8/9 bp. To determine the lengths of recognition sites of IS50 we mutagenized IS50 in vitro and tested the transposition proficiency of the resulting mutants. Our results show that the recognition sites at each end of IS50 are about 19 bp long. These findings suggest models for the evolution of IS elements from simpler immobile gene complexes.

The terminal nucleotides of retrovirus DNA are required for integration but not virus production

Deletion of specific nucleotides at either end of the long terminal repeat of the avian retrovirus, spleen necrosis virus, results in replication-competent but integration-defective virus. This result supports two conclusions: (1) the 5-base pair terminal inverted repeats and three to seven adjacent nucleotides are required for integration; (2) integration of retrovirus DNA is not required for retrovirus gene expression.

DNA sequences at the ends of transposon Tn5 required for transposition

Transposons are a class of genetic elements that can move from one site in a cell's genome to another independently of the cell's general recombination system. Little is known about the mechanism of transposition of compound transposons such as Tn5, but it is thought that a transposon-encoded protein (a transposase) must recognize the outer ends of the element and, together with host factors, catalyse the transfer of the internal DNA into a new site in a manner that may involve replication. It has previously been shown that the synthesis of an IS50R-encoded protein (protein 1) is an essential requirement for Tn5 transposition. Here we demonstrate that a structure containing only the outer 186 base pairs (bp) of both inverted repeats is capable of being efficiently complemented to transpose in Escherichia coli, provided IS50R is located close by on the same replicon. In addition, Bal31-generated deletions indicate that 16-18 bp of the outer end of IS50L are required for transposition. This 16-18-bp sequence contains the 8-9-bp small inverted repeat present at each end of IS50 plus a 9-bp sequence which is homologous to an interrelated sequence present in four copies in the chromosomal origin of replication in a variety of Gram-negative bacteria. This sequence organization suggests that the ends of Tn5 may function to provide a recognition site for the Tn5 transposase adjacent to a sequence recognized by the host replication system.

Postexcision transposition of the transposon Tn10 in Escherichia coli K12

An experimental analysis of the fate of transposon Tn10 after excision from a proA::Tn10 site localized on the plasmid F' leads to the conclusions: 1. The precise excision is a progressive process. Its probability is estimated per time unit. 2. An excised Tn10 is always integrated into a different genetic locus. 2. An excised Tn10 is always integrated into a different genetic locus. 3. The kinetics of postexcision transposition are sometimes very slow. The excised transposon is inherited in one cell line in spite of cell multiplication. 4. The processes of excision and secondary insertion have no absolute requirement for the recA+ genotype but they are strongly enhanced in recA+ cells. 5. The kinetics of postexcision transposition are strongly dependent on the genetic site from which the transposon was excised. 6. The probability of postexcision transposition is fully determined by the probability of excision and depends on the genotype of the host and many other factors.

A product of the TN5 transposase gene inhibits transposition

The bacterial transposon Tn5 possesses a regulatory mechanism that allows it to move with higher efficiency when it is first introduced into a cell than after it is established. Tn5 is a composite transposable element containing inverted repeats of two nearly identical elements, IS50R, which encodes the transposase protein necessary for Tn5 movement, and IS50L which contains an ochre mutant allele of the transposase gene. Data presented here show that Tn5 transposition is inhibited about 50-fold in cells of Escherichia coli which already carry IS50R in the multicopy plasmid pBR322. If the cells contain a plasmid carrying either IS50L instead of IS50R, or derivatives of IS50R in which the transposase gene has been mutated, little if any inhibition of Tn5 transposition is found. Although inhibition had previously been hypothesized to require interaction between the products of IS50L and IS50R, our results show that IS50R alone is sufficient to mediate inhibition and suggest that the inhibitor is a product of the transposase gene itself.

Evolution of transposons: natural selection for Tn5 in Escherichia coli K12

A novel in vivo effect of the transposable element Tn5 has been observed in chemostats when certain isogenic Tn5 and non-Tn5 strains of Escherichia coli compete for a limiting carbon source in the absence of kanamycin. The Tn5-bearing strain has a more rapid growth rate and increases in frequency from 50% to 90% within the first 15 to 20 generations. The effect occurs when Tn5 is inserted at a variety of chromosomal locations or when the element is carried by an episome, but it is strain specific, having been observed in two out of three strains examined. (For reasons unknown, the effect has not been observed with derivatives of strain CSH12.) Although the growth-rate advantage of Tn5 is independent of nutrient concentration and generation time, it can be reduced by prior adaptation of the strains to limiting conditions, and the amount of reduction is proportional to the length of prior adaptation. The growth-rate effect is evidently not caused by beneficial mutations induced by Tn5 transposition, as Tn5-bearing strains selected in chemostats retain their initial Tn5 position and copy number. However, the effect does not occur in Tn5-112, a transpositionless deletion mutation missing the transposase-coding region of the right-hand IS sequence flanking the element. Since Tn5-112 retains a functional kanamycin-phosphotransferase gene, this gene is not responsible for the growth-rate effect. Thus, the effect evidently requires transposase function, but it does not involve actual transposition of the intact element. Altogether, these data provide a mechanism for the maintenance of Tn5 in bacterial populations in the absence of kanamycin, and they suggest a model for the proliferation and the maintenance of IS sequences and transposable elements in the absence of other identifiable selection pressures.

Structural requirement for IS50-mediated gene transposition

Replicative transposition is signaled by the formation of cointegrates in which donor and target replicons are joined by direct repeats of a transposable element. Elements not generating such cointegrates may move by a conservative breaking and joining process. The IS50 elements forming the terminal repeats of Tn5 [which carries the determinant for kanamycin resistance (Kanr)] contain genes and sites necessary for transposition and mediate the movement of any DNA segment they bracket. To determine if IS50 generates cointegrates, the products of transposition from pBR322::Tn5 plasmids to an F factor in recA-Escherichia coli were examined. With monomeric pBR322::Tn5 plasmids, transposition of Kanr (from Tn5) was generally not accompanied by movement of the determinant for ampicillin resistance (Ampr) (from the pBR322 vector). With dimeric pBR322::Tn5 plasmids, by contrast, half of the transpositions of kanr were accompanied by transposition of ampr. Restriction endonuclease analyses indicated that these F-Kanr Ampr chimeras contained inserts of a single copy of the pBR322 vector sequence bracketed by one Tn5 element and one IS50 element or by a pair of Tn5 elements. None of 79 chimeras tested was a true cointegrate. Because IS50 seems to move only a segment of the donor replicon it is proposed that IS50 transposition is conservative.

Control of transposon Tn5 transposition in Escherichia coli

Tn5 is a composite transposable element in which the insertion sequences IS50R and IS50L bracket a central region encoding kanamycin resistance (kanr). IS50R encodes a functional transposase, whereas IS50L contains the promoter of the kanr gene. To determine the relative activities of IS50R and IS50L in transposition we examined the structures of chimeric DNA molecules generated by insertion of segments of pBR322::Tn5 dimeric plasmids into red- lambda phage in recA- Escherichia coli. Restriction endonuclease analyses showed that the inserted sequences contained direct terminal repeats of pairs of IS50R or of IS50L elements and that the frequencies of usage of IS50R vs. IS50L depended on the position and orientation of Tn5 in the plasmid vector: IS50R was used preferentially when Tn5 was in transcriptionally quiescent regions of the vector (in either orientation) or when IS50L was immediately downstream from a strong promoter in the vector. In contrast, IS50L was used preferentially when IS50R was downstream from a strong promoter. We conclude IS50R tends to be used preferentially but that when transcription impinges on the end of an IS50 element the participation of that element in transposition is inhibited.