New ultrarare restriction site-carrying transposons for bacterial genomics

Pathoadaptive mutations that enhance virulence: genetic organization of the cadA regions of Shigella spp

Pathoadaptive mutations improve the fitness of pathogenic species by modification of traits that interfere with factors (virulence and ancestral) required for survival in host tissues. A demonstrated pathoadaptive mutation is the loss of lysine decarboxylase (LDC) expression in Shigella species that have evolved from LDC-expressing Escherichia coli. Previous studies demonstrated that the product of LDC activity, cadaverine, blocks the action of Shigella enterotoxins and that the gene encoding LDC, cadA, was abolished by large chromosomal deletions in each Shigella species. To better understand the nature and evolution of these pathoadaptive mutations, remnants of the cad region were sequenced from the four Shigella species. These analyses reveal novel gene arrangements in this region of the pathogens' chromosomes. Insertion sequences, a phage genome, and/or loci from different positions on the ancestral E. coli chromosome displaced the cadA locus to form distinct genetic linkages that are unique to each Shigella species. Hybridization studies, using an E. coli K-12 microarray, indicated that the genes displaced to form the novel linkages still remain in the Shigella genomes. None of these novel gene arrangements were observed in representatives of all E. coli phylogenies. Collectively, these observations indicate that inactivation of the cadA antivirulence gene occurred independently in each Shigella species. The convergent evolution of these pathoadaptive mutations demonstrates that, following evolution from commensal E. coli, strong pressures in host tissues selected Shigella clones with increased fitness and virulence through the loss of an ancestral trait (LDC). These observations strongly support the role of pathoadaptive mutation as an important pathway in the evolution of pathogenic organisms.

The form of chromosomal DNA molecules in bacterial cells

The circular concept of the bacterial chromosome was based initially on experiments involving conjugation mapping and autoradiographic imaging of DNA. This view was then supported by DNA fragment mapping, genome sequencing, and the analysis of linear DNA produced by a single cleavage of chromosomal DNA. A circular chromosome is also indicated by the existence of a mechanism for segregating dimeric chromosomes produced by recombination and the replication of DNA on both sides of the replication terminus. The evidence for circularity is reviewed here and found to be compatible with either a circular or a linear chromosomal DNA molecule. Moving pictures of ethidium-stained DNA revealed most chromosomal DNA as a rosette form with loops emanating from a dense node or as a network of strands lacking a node. This description applies to Escherichia coli, Agrobacterium tumefaciens, Pyrococcus endeavorii, Vibrio cholerae, and both the linear-mapping chromosome of Streptomyces lividans and its circular-mapping derivative. Networks without nodes were found for two linear-mapping Borrelia species. For the E. coli chromosome, open-form circles of various sizes were found only at extremely low frequency. The node of the rosette was reduced in size or eliminated in recA mutants, as well as by treatment with either ribonuclease, topoisomerase IV, 1 M NaCl, or lysozyme. A model is presented for the bacterial chromosome in which the DNA is compacted by many points of strand association (including recombination junctions, tangles and knots) created during the repair of DNA damage that occurs many times in each chromosome replication cycle.

Integrated genomic map from uropathogenic Escherichia coli J96

Escherichia coli J96 is a uropathogen having both broad similarities to and striking differences from nonpathogenic, laboratory E. coli K-12. Strain J96 contains three large (>100-kb) unique genomic segments integrated on the chromosome; two are recognized as pathogenicity islands containing urovirulence genes. Additionally, the strain possesses a fourth smaller accessory segment of 28 kb and two deletions relative to strain K-12. We report an integrated physical and genetic map of the 5,120-kb J96 genome. The chromosome contains 26 NotI, 13 BlnI, and 7 I-CeuI macrorestriction sites. Macrorestriction mapping was rapidly accomplished by a novel transposon-based procedure: analysis of modified minitransposon insertions served to align the overlapping macrorestriction fragments generated by three different enzymes (each sharing a common cleavage site within the insert), thus integrating the three different digestion patterns and ordering the fragments. The resulting map, generated from a total of 54 mini-Tn10 insertions, was supplemented with auxanography and Southern analysis to indicate the positions of insertionally disrupted aminosynthetic genes and cloned virulence genes, respectively. Thus, it contains not only physical, macrorestriction landmarks but also the loci for eight housekeeping genes shared with strain K-12 and eight acknowledged urovirulence genes; the latter confirmed clustering of virulence genes at the large unique accessory chromosomal segments. The 115-kb J96 plasmid was resolved by pulsed-field gel electrophoresis in NotI digests. However, because the plasmid lacks restriction sites for the enzymes BlnI and I-CeuI, it was visualized in BlnI and I-CeuI digests only of derivatives carrying plasmid inserts artificially introducing these sites. Owing to an I-SceI site on the transposon, the plasmid could also be visualized and sized from plasmid insertion mutants after digestion with this enzyme. The insertional strains generated in construction of the integrated genomic map provide useful physical and genetic markers for further characterization of the J96 genome.

Targeting and retrofitting pre-existing libraries of transposon insertions with FRT and oriV elements for in-vivo generation of large quantities of any genomic fragment

A procedure is described that converts the pre-existing transposon insertion libraries to a collection of 'pop-out' strains, each allowing generation of 20- to 100-kb genomic fragments directly from the genome. The procedure consists of two steps: (1) single transposon insertions are targeted and retrofitted with excision and amplification elements (FRT and oriV), by homologous recombination with an FRT-oriV-carrying plasmid; and (2) two retrofitted neighbouring transposons are brought together by P1 transduction. From each strain, a 20- to 100-kb genomic fragment, bound by a pair of retrofitted transposons, could be excised and amplified upon supplying in trans the excision (Flp) and replication (TrfA) functions. To enhance the efficiency of crossing-in the FRT-oriV cassette, we transiently increased the copy number of our retrofitting plasmids using a temperature-sensitive TrfA-supplying helper plasmid. Using FRT-oriV and helper plasmids, we retrofitted four Tn10KmR and three Tn10CmR insertions. Subsequently, the FRT-oriV retrofitted insertions were crossed with each other in pairs (KmRxCmR), using P1 phage transductions. The resulting CmRFRT-[28-65-kb]-KmRFRT strains were transformed with a plasmid expressing FLP and trfA genes from the tightly controlled Ptet promoter. Induction of this tightly repressed promoter by autoclaved chlortetracycline (cTc) resulted in the efficient excision and amplification of genomic fragments located between FRT sites, but only in productive strains, i.e. having two parallel FRTs. We have shown that genomic fragments of 28-, 40-, 50- and 65-kb were efficiently excised and amplified. Furthermore, we could convert non-productive strains (having FRTs in non-parallel orientation), to productive combination of parallel FRTs, because one of the FRT elements was flanked by two convergent loxP sites, and thus could be inverted by the Cre function delivered either by the P1 phage or by a specially constructed temperature-sensitive Plac-cre plasmid. Although several microbial genomes were recently sequenced, the described method will help in supplying large quantities of any genomic fragment (prepared without the conventional cloning and its artifacts) for refined sequence comparison among strains and species, and for further analysis of uncharacterized ORFs, various mutations, and regulatory elements or functions. The excised and circularized DNA fragments (plasmids) could be propagated like any other large plasmids but only in hosts that could supply the appropriate Rep function. Our original 'pop-out' method [Pósfai et al. (1994) Nucleic Acids Res. 22, 2392-2398] was already employed for sequencing of the E. coli genome [Blattner et al. (1997) Science 277, 1453-1462]. Moreover, the Flp-mediated recombination between two FRT elements resulted in bacterial strains with large deletions (for parallel FRT orientations) or with large inversions (for inverted FRT orientations).

Subdivision of the Escherichia coli K-12 genome for sequencing: manipulation and DNA sequence of transposable elements introducing unique restriction sites

A transposon-based method of introducing unique restriction sites was used for subdivision of the Escherichia coli genome into a contiguous series of large non-overlapping segments spanning 2.5Mb. The segments, sizes ranging from 150 to 250kb, were isolated from the chromosome using the inserted restriction sites and shotgun cloned into an M13 vector for DNA sequencing. These shotgun sizes proved easily manageable, allowing the genomic sequence of E. coli to be completed more efficiently and rapidly than was possible by previously available methods. The 9bp duplication generated during transposition was used as a tag for accurate splicing of the segments; no further sequence redundancy at the junction sites was needed. The system is applicable to larger genomes even if they are not already well-characterized. We present the technology for segment sequencing, results of applying this method to E. coli, and the sequences of the transposon cassettes.

Integrated physical and genetic mapping of Bacillus cereus and other gram-positive bacteria based on IS231A transposition vectors

The genome structure of Bacillus cereus is relatively complex, its DNA being modulated between a size-varying chromosome and large plasmids. To study the genetic organization of the B. cereus type strain ATCC 14579, thermosensitive transposition vectors were designed on the basis of IS231A-derived cassettes containing uncommon restriction sites. A highly preferred insertion site for IS231A was detected in the chromosome by Southern blotting and pulsed-field gel electrophoresis (PFGE) analyses of independent insertion mutants. However, once this insertional hot spot was occupied, secondary IS231A insertions occurred randomly, as demonstrated by isolation of independent B. cereus auxotrophs at a frequency of approximately 0.6%. The hot-spot site, as well as several auxotrophic mutations, were mapped by using NotI, SfiI, and AscI PFGE restriction profiles. It was confirmed by sequencing that one of the insertions, generating an Ade- phenotype, had disrupted a gene of the purine synthesis pathway. These results showed that combined PFGE and sequencing analyses of mini-IS231A insertions enable the construction of integrated physical and genetic maps of B. cereus type strain. Moreover, the presence of the ultrarare I-SceI restriction site in the mini-IS231A allowed the isolation, in double-insertion mutants, of contiguous and nonoverlapping large chromosomal fragments, convenient for direct sequencing. The system detailed in this report is therefore a powerful tool for comparative genetic studies among members of the B. cereus group (i.e., B. cereus, B. thuringiensis, B. mycoides, and B. anthracis) and could also be applied to more distantly related gram-positive bacteria.

"Black holes" and bacterial pathogenicity: a large genomic deletion that enhances the virulence of Shigella spp. and enteroinvasive Escherichia coli

Plasmids, bacteriophages, and pathogenicity islands are genomic additions that contribute to the evolution of bacterial pathogens. For example, Shigella spp., the causative agents of bacillary dysentery, differ from the closely related commensal Escherichia coli in the presence of a plasmid in Shigella that encodes virulence functions. However, pathogenic bacteria also may lack properties that are characteristic of nonpathogens. Lysine decarboxylase (LDC) activity is present in approximately 90% of E. coli strains but is uniformly absent in Shigella strains. When the gene for LDC, cadA, was introduced into Shigella flexneri 2a, virulence became attenuated, and enterotoxin activity was inhibited greatly. The enterotoxin inhibitor was identified as cadaverine, a product of the reaction catalyzed by LDC. Comparison of the S. flexneri 2a and laboratory E. coli K-12 genomes in the region of cadA revealed a large deletion in Shigella. Representative strains of Shigella spp. and enteroinvasive E. coli displayed similar deletions of cadA. Our results suggest that, as Shigella spp. evolved from E. coli to become pathogens, they not only acquired virulence genes on a plasmid but also shed genes via deletions. The formation of these "black holes," deletions of genes that are detrimental to a pathogenic lifestyle, provides an evolutionary pathway that enables a pathogen to enhance virulence. Furthermore, the demonstration that cadaverine can inhibit enterotoxin activity may lead to more general models about toxin activity or entry into cells and suggests an avenue for antitoxin therapy. Thus, understanding the role of black holes in pathogen evolution may yield clues to new treatments of infectious diseases.

Genomic fluidity of Bordetella pertussis assessed by a new method for chromosomal mapping

The genomic organization of Bordetella pertussis strains has been examined by using a new method. This method does not depend on the prior determination of a restriction map of the bacterial chromosome but is based on the ability to measure directly the distance between two genes. This is accomplished through the integration at each gene of a suicide vector containing a cleavage site for the intron-encoded endonuclease I-SceI, which is not otherwise found in the chromosome. Integration is mediated by homologous recombination between the chromosomal and cloned plasmid copies of a gene of interest. Digestion with I-SceI gives rise to a fragment the size of which represents the distance between the two genes. Multiple pairwise determinations within a set of genes provide sufficient information to derive a map of the relative gene positions. Mapping a set of 11 to 13 genes for five strains of B. pertussis and one strain of B. parapertussis revealed extensive divergence of gene order between B. pertussis Tohama I, B. pertussis 18-323, and B. parapertussis ATCC 15311. Less extensive divergence of gene order was observed between B. pertussis Tohama I and B. pertussis Tohama III, BP165, and Wellcome 28, with most of the observed differences explainable by large inversions.

Homing endonucleases: keeping the house in order

Homing endonucleases are rare-cutting enzymes encoded by introns and inteins. They have striking structural and functional properties that distinguish them from restriction enzymes. Nomenclature conventions analogous to those for restriction enzymes have been developed for the homing endonucleases. Recent progress in understanding the structure and function of the four families of homing enzymes is reviewed. Of particular interest are the first reported structures of homing endonucleases of the LAGLIDADG family. The exploitation of the homing enzymes in genome analysis and recombination research is also summarized. Finally, the evolution of homing endonucleases is considered, both at the structure-function level and in terms of their persistence in widely divergent biological systems.