The Bacterial Chromosome

The challenge to those trying to explain chromosome behavior is to identify the critical sequence elements. One might expect that a genome would be under constant selective pressure to expand, adding new genes that broaden the capabilities of the organism. Three aspects of population biology and lifestyle limit this expansion: mutation rate, population size, and recombination frequency. First, if all other factors remain constant, mutation rate limits genome size; beyond a critical point, the genome can expand only if the mutation rate drops. Second, larger populations allow genome expansion because selection is more effective and genes with a smaller fitness contribution can be maintained. Third, sexual recombination allows genome expansion by permitting assembly of intact information sets from those damaged by mutation. Selfish mechanisms used by phage, transposons, and plasmids can also contribute to maintenance of genes that are respectable chromosome residents. The chapter describes some genome features that depend to varying extents on their selfish behavior for their maintenance in the genome. These assumptions can be used in reverse to draw conclusions regarding the fitness contribution of genes based on their genomic position. Two short sequence elements have been extensively studied with regard to their effects on chromosome replication (Ter) and recombination (Chi). In Escherichia coli and Salmonella enterica, the positions and orientation of seven repeated rRNA operons are conserved. Several small sequence elements are found in the genome of E. coli.

The model of a static bacterial chromosome arose from early comparisons of the genetic maps of Escherichia coli and Salmonella typhimurium. Analyses of complete genome sequences by several methods revealed that the differences in gene content were the result of two complementary processes: the gain of new genes by horizontal gene transfer from distantly related organisms, and the loss of ancestral genes from descendent lineages. Directional mutation pressures provide a distinct ‘‘fingerprint’’ to a bacterial genome owing to the differential mutational proclivities of DNA polymerases, the nature and number of mismatch correction systems, the numbers and abundances of tRNA species, and even relative concentrations of precursor nucleotide pools. Thus, genes which appear atypical in their current genomic context may reflect the direction pressures of a donor genome. Aside from changes in gene content, gene order has also been found to be more plastic than once assumed. Mechanisms for DNA rearrangement are well known and have been well measured in the laboratory. Yet despite the opportunities for chromosomal rearrangement, the genetic maps of E. coli and S. enterica seemed to be largely congruent, save the inversion about the terminus of replication. The genome, with all its dynamic parts, steers the organism into the environmental space it is best suited to exploit. Rather than a stale collection of genes having reached optimal performance after billions of years of evolution, one may view a bacterial genome as an ever-changing consortium of genes which cooperate in perpetuating their host organism.

This chapter focuses on the doublestranded DNA (dsDNA) phages, and especially on the temperate phages. While virulent phages certainly perform transduction and engage in evolutionary sparring with their hosts and so influence their evolution, the chapter focuses mainly on the complex interactions of temperate phages with their hosts. Bacteriophages may thus have contributed to the current compact nature of bacterial genomes. The approximately 100 currently published bacterial genome complete nucleotide sequences, and about 285 prophages are related to known bacteriophages. Of the more than 280 prophages in the currently sequenced bacterial genomes, only a few are known to be fully functional bacteriophages. There are two rather complex types of genetic entity in which this appears to have happened: the phage tail-like bacteriocins and the gene transfer agents. To date, protection from other phages and disease virulence factors are the lysogenic conversion genes that have been discovered and studied in the laboratory, but this likely reflects their ease of study and the lifestyles of the hosts studied. Our ideas about how bacteriophages have affected the nature of the bacterial chromosome are necessarily based on extrapolations from things we know about bacteriophage biology and from inferences based on the current structure of the bacterial genomes, and not on direct observation of those processes over evolutionary time.

New technologies made possible by this sequence data, such as DNA microarrays, in combination with the small size and ease of genetic manipulation of bacteria, now make it possible to identify the complete genetic regulatory circuitry that controls the bacterial cell. Analysis of the global gene expression profile of the bacterial cell during its cell cycle, under conditions of environmental challenge, and during pathogen invasion of host organisms will provide an unprecedented understanding of the bacterial cell as an integrated system. This chapter addresses the use of microarrays for study of a wide range of microbiological problems with emphasis on the profoundly different results that this genome-wide technique provides relative to the analysis of single genes and conventional forward genetics. By assaying the response of all genes to a given genetic or environmental perturbation in parallel and simultaneously, the microarray results identify whole pathways or subroutines of the organism’s genetic regulatory circuitry. The immobilized arrays, or spots, of DNA are typically the products of PCR that generate amplicons ranging from a few hundred base pairs to several kilobases in length. Application of microarray-based genomic analysis to study the cell cycle of Caulobacter crescentus, has recently led to a dramatic increase in one's understanding of regulation of the bacterial cell cycle. Although microarrays were initially developed to analyze RNA levels, they can also be used to examine DNA samples.

This chapter focuses on the major nucleoid-associated proteins and summarizes one's current knowledge of how these proteins contribute to the structure of the nucleoid and function in specific reactions involving the chromosome. The current view is that the Escherichia coli nucleoid is dominated by five different proteins: HU, integration host factor (IHF), H-NS, StpA and Fis under nutrient-rich exponential growth conditions. HU mutant cells display the most severe and varied phenotypes in comparison to strains containing mutations in the other major nucleoid proteins. Supercoiling of chromosomal and plasmid DNA is partially relaxed in mutants lacking HU, supporting an in vivo role for HU in regulating the degree of supercoiling. HU plays several interconnected roles in DNA replication and appears to be involved in several steps involving initiation of chromosomal DNA synthesis, chromosomal partitioning, and cell division. The genes that were differentially expressed were examined for putative IHF binding sites in the 500 bp upstream of the transcription initiation site. The criteria for identifying high-affinity sites were that the sites must have a 12 out of 13 bp match with the core consensus sequence and that at least 10 of the 15 bp 5' (upstream) to the core consensus must be dA-dT base pairs. The study identified 46 candidates that matched the criteria, some of which were documented in other studies. The study also identified seven genes, including ihfA, that were expressed only when IHF was present and eight genes that were expressed only in IHF-deficient cells.

This chapter reviews classical results and incorporates a selected set of facts into models of chromosome dynamics. It emphasizes the topology of DNA during replication and transcription because these two processes exert the most dramatic topological influence on DNA. One focus is on structural properties of bacterial chromosomes that allow rapid adaptability. The chapter also discusses chromosome structure considered from a stochastic as opposed to a highly ordered perspective. The introduction of a limited number of single-strand nicks causes nucleoid sedimentation values to decrease in gradual stages. In vitro nucleoid studies stimulated searches for the controlling elements of domain behavior. Three generalizations of the his-cob interval have been extended to 2% of the chromosome. First, supercoil domains are more abundant when cells are undergoing DNA replication than when DNA replication is suppressed. Second, the resolution efficiency diminishes as a first-order function of distance along the chromosome. Third, the number of domains detected depends on the time period of the assay. Global structure provided by four topoisomerases and a group of about 10 DNA-binding proteins allows a plasticity that is useful for adaptation to different physiological conditions. These observations indicate that the organizational framework of a bacterial chromosome must be described in statistical terms rather than the highly ordered and predictable models that represent the crystal structures of many folded proteins that do biochemical work on the chromosome.

One of the most remarkable features of prokaryotes is their ability to remain viable for very long periods under conditions that are not propitious for growth. The intrinsic chemical and physical vulnerability of DNA molecules and the lethal effects caused by unrepaired DNA lesions, even if they occur at low frequency, highlight the need for particularly efficient DNA protection mechanisms. The chapter presents a concise survey of data derived from light and electron microscopy techniques on the architecture of the chromosome in actively growing bacteria. It proceeds to briefly describe phase transitions that characterize DNA molecules. The chapter considers these two issues a prerequisite for a deeper understanding of the factors that determine the structure of chromatin in stationarystate bacteria. The organization of the chromosome within bacterial cells and the factors that dictate and modulate this organization remain less thoroughly understood than the structure of eukaryotic chromosomes. Several factors conspired to bring about this situation. These include the small size of the prokaryotic cell that restricts the effectiveness of light microscopy, the relatively low content of DNA-binding proteins associated to the bacterial chromosome which resulted in severe artifacts in conventional electron microscopic studies, as well as the apparent absence of structural order and architectural hierarchy within the nucleoid. Although this survey is concerned with stationary-phase chromosomes, a sideways glance at sporulating bacteria seems to be in place.

In eukaryotes, DNA replication appears to occur in stationary factories that can be visualized as foci in the nucleus via either immunofluorescence microscopy of replication proteins or fluorescence microscopy of live cells carrying green fluorescent protein-tagged replication proteins. In bacteria, the division plane at midcell is a crucial locus where most of the events leading to cytokinesis take place. Using green fluorescent proteintagged subunits of the Pol III HE, researches have shown that the DNA polymerases assemble and replicate DNA at midcell in Bacillus subtilis. Studies on the effect of moving dif on chromosomal dimer resolution and on the tolerance of deletions and inversions in the terminus region have led to the suggestion that the chromosome is polarized, meaning that the directionality of the path taken along the chromosome away from oriC has some physical manifestation. The proteins required for restart primosome assembly—PriA, PriB, PriC, DnaT, DnaB, DnaC, and DnaG—were discovered originally because they acted to assemble a primosome on φX174 viral DNA during its conversion to the replicative form. In Escherichia coli, estimates of the probability that a replication fork formed at oriC will be arrested before completing replication of one arm of the chromosome vary from 0.15 to almost certainty.

The complex process of bacterial chromosomal replication can be divided into several stages: initiation, priming of chain starts, chain elongation, and termination. Since much of what is known about the initiation of bacterial chromosomal replication comes from studies of Escherichia coli, this chapter concentrates on initiation in that organism. The roles of the crucial sequence elements in the initiation and regulation of chromosomal replication are discussed in this chapter. DnaA protein, a sequence-specific DNA-binding protein, is responsible for setting in motion the cascade of events for initiating chromosomal replication, including origin recognition, strand opening, and loading of the replicative helicase at the sites of the future bidirectional replication forks. SeqA tetramers must interact properly and form active aggregates for binding to hemimethylated DNA to occur. This SeqA aggregation may be important not only for regulating chromosomal replication, but also for chromosomal segregation. Bypassing the normal, DnaA-dependent initiation of chromosomal replication from oriC via constitutive stable DNA replication relieves the growth arrest of cells lacking sufficient acidic phospholipids. Regulated initiation of chromosomal replication likely involves not only its timing during the cell cycle, but also where it happens within the cell. Recently, significant advances have been made in our knowledge of the initiation of chromosomal replication.

This chapter describes the structure and mechanisms of action of bacterial DNA replication proteins, and how their activities are coordinated for efficient duplication of chromosomal DNA. The proteins holoenzyme/replicase function together as catalytic or structural components of the chromosomal DNA replication machinery known as a replisome. At the heart of the replisome is the replicase, or DNA polymerase holoenzyme-itself a complex protein machine comprising DNA polymerase and the accessory clamp and clamp loader proteins. Prokaryotic single-stranded DNA-binding protein (SSB) and eukaryotic replication protein A (RPA) are known to bind several proteins involved in DNA metabolism. The interactions between Escherichia coli SSB and primase and between SSB and χ protein of γ/τ complex are critical for assembly of DNA polymerase onto primed DNA. Each replisomal protein provides a specific function or interaction at a specific point in the DNA replication pathway, enabling highly efficient DNA synthesis. Cell survival depends on the ability of the replisomal protein machinery to overcome such blocks in the path of DNA elongation. Bypass polymerase activity appears to be distributive, suggesting that once the polymerase traverses the lesion it "falls off," allowing reassembly of the replicative DNA polymerase and continued DNA elongation. The processes described clearly indicate that the pathways of DNA replication, recombination, and repair are intimately intertwined.

The membrane affinity of the SeqA protein tends to make it insoluble in whole-cell extracts unless the salt concentration is elevated. If SeqA protein binding forms a coherent filament of protein on the DNA with the sequences intervening between the GATC sites looped out, the newly replicated DNA would be organized and compacted as it emerges from the replication fork. The E. coli chromosome replication forks are not thought to move about the nucleoid as they progress around the chromosome. Chromosome segregation is a direct consequence of replication and occurs concomitantly with it. The DNA replication process itself drives DNA segregation, pushing the newly replicated DNA outward from the anchored replication forks toward opposite cell poles. The properties of the SeqA protein and its selective binding to the newly replicated DNA at the replication forks suggest that it might be directly involved in some or all of these processes. If the single-strand contact persisted through replication, newly replicated DNA would be uniquely marked by having protein subunits on one strand only. If this ‘‘hemidecorated’’ DNA were selfassociating, a mechanism similar to that proposed for SeqA might operate in the absence of any methylation signals.

This chapter highlights the current knowledge of replication restart in Escherichia coli and T4-infected E. coli. T4 uses homologous recombination to initiate most of its DNA replication. Soon after phage gene expression commences, DNA replication is initiated from internal replication origins, and the forks so generated travel toward the ends of the infecting genome. UvsW is an RNA-DNA helicase that disrupts the origin R-loop, an essential intermediate in origin replication. Many rounds of recombination-dependent DNA replication (RDR) convert the intracellular form of T4 DNA into a long concatemer. Investigations of phage T4 hotspots for marker rescue recombination provided a dramatic demonstration of the coupling between replication, DNA damage, and recombination. The hotspots were first detected as regions of the genome where genetic markers could be rescued by homologous recombination from UV-irradiated phage at an inflated frequency. The coordinated action of several proteins-PriA, PriB, PriC, DnaT, and DnaC--is required for replication reinitiation at nonorigin sequences in E. coli. PriA is the key protein in assembly of the primosome complex, the base of the scaffold that leads to the loading of DnaB. In addition to its primosome assembly activity, PriA has an ssDNA-dependent ATPase activity, which is also dependent on primosome assembly site (PAS) sites in the presence of single-strand binding protein (SSB). RecG strongly prefers forks with a single-strand gap on the leading strand; this and other results argue for a special role of RecG in the replication of damaged DNA.

Research on termination of the replication cycle and the terminus region was prompted by the discovery of bidirectional replication of the Escherichia coli chromosome. Studies of the Tus-Ter system and the terminus region had other fruitful consequences, since they led to the recognition of phenomena important in the cell cycle, which are restricted to the terminus region by features of the global organization of the bacterial nucleoid. This chapter describes one's current understanding of the features of the terminus region. The core of this chapter deals with resolution of chromosome dimers by the dif-XerCD/FtsK system and its regional control. Other aspects, such as the organization in nondivisible zones (NDZ), local hyper recombination, and evolution of terminus sequences, are also presented in this chapter. Stress is placed the role of the terminus in the ultimate operations of the replication cycle and the postreplicative processing of sister chromosomes. The first evidence for bipolarization of the terminus region was provided by the distribution of the Ter pause sites. According to the author, the arrangement of polar pause sites on the chromosome, which forces replication to terminate in the center of the terminus macrodomain near dif, facilitates temporal harmony between end of replication and triggering of events which, like movements of the TER domains away from the septal plane, involve the chromosome polarization centered on dif. The function of the terminus region about which the most is known is the resolution of circular dimer chromosomes.

Bacterial RNA polymerase provides the central model for the transcription elongation complex and its various interesting fates-backtracking and correction by Gre protein-mediated transcript cleavage, transcription termination, and the antitermination controls that were discovered in bacteria. RNA polymerase and its transcription factors have functions beyond their obvious activity to provide RNA molecules to the cell, reflecting the fact that RNA polymerase and the process of transcription must have evolved as DNA arose from the primal RNA world-neither is worth much without the other. There is evidence or informed speculation implicating RNA polymerase and transcription proteins in processes of replication, DNA repair, and cell division. Thus, transcription by RNA polymerase activates the origins of replication of Escherichia coli and phage λ in some structural way independent of the RNA product. Just as transcription and replication coevolved, so did the coordination of chromosome segregation and cell division arise in the context of both. DNA is transcribed as it moves about the cell in an organized fashion during replication. RNA is translated at the same time, causing an added complication when emerging membrane proteins are inserted into the membrane and provide points of fixation for the complex.

This chapter describes crystal structures of RNA polymerase (RNAP) structures and their implications for understanding the mechanism of transcription and the regulation of key steps in the transcription cycle. There are three main steps in the transcription cycle: initiation, elongation, and termination. The transcription elongation complex (TEC) is processive and extremely stable, transcribing at an average rate of 30 to 100 nucleotides for tens of kilobases down the DNA template. Much is known about the general architecture of RNAP and the nucleic acid scaffold in the TEC from biochemical experiments. The chapter provides a comprehensive overview of what has been learned from the bacterial RNAP structures and models. Rifampin positioned in binding pocket would block growth of the RNA chain past two or three nucleotides, explaining the bactericidal effect of the antibiotic. In the holoenzyme structure, two rearrangements in σ region 2 are evident. First, a loop that covers the core binding surface in region 2.2 moves out of the way. Second, the bundle of helices made up of regions 1.2 and 2.1 to 2.4 rotates about 12⁰ relative to the nonconserved region. The single-stranded RNA transcript in the RNA-exit channel has also been shown to be important for the stability of the elongation complex, possibly due to the flap domain closing down around the RNA. The publication of crystal structures for multiple forms of RNAP has made possible a much more detailed examination of the function of the enzyme and the mechanisms of catalysis, promoter recognition, and transcriptional activation.

Most activators of transcription initiation that affect s70-containing RNA polymerase (RNAP) are sequence-specific DNA-binding proteins that bind to recognition sites located upstream of the core promoter. The chapter explains the mechanism of action of one of the most thoroughly characterized activators of σ70-dependent transcription in Escherichia coli, the cyclic AMP receptor protein (CRP), also known as the catabolite activator protein. A possible explanation for the change in the kinetics of λcI-dependent activation is discussed in the chapter. The chapter talks about activators that bind to DNA and affect the process of transcription initiation by making direct contacts with RNAP. The detailed structural basis for this activator induced promoter remodeling has recently been revealed by the crystal structure of the Bacillus subtilis BmrR protein, in complex with promoter DNA and a drug cofactor. In general, transcription activators that bind DNA and contact RNAP are thought to bind their specific DNA recognition sites and then, once appropriately positioned on the DNA, to interact with RNAP. The chapter discusses examples of activators that work only when bound to specific sites on the DNA, an activator that must be tethered to the DNA but remain mobile, and activators that can work directly from solution.

This chapter provides a review of the current structural and kinetic models for transcription elongation and termination. It describes the regulatory molecules that are known to influence the elongation/termination decision by RNA polymerase (RNAP), with the emphasis on the most recent findings and on the mechanism of ‘‘active’’ regulators whose actions are not limited to changes in RNA folding. Interactions between RNAP and the nucleic acid chains, as well as the RNA:DNA pairing in the hybrid, all contribute to the extraordinary stability of the elongating transcription elongation complexes (TECs). RNA release is triggered at sites where the nascent RNA folds into a stable, GC-rich hairpin followed by a stretch of the U-rich RNA. Rho is the main termination protein in Escherichia coli, where it is thought to control ~ 50% of all termination events. Feedback control of the operons that encode ribosomal protein synthesis is commonly accomplished by autogenous regulation by one of the products. Alc protein terminates transcription at several sites on a nonmodified host DNA, but not on the phage DNA that contains hydroxymethyl cytosine residue. The bgl operon in E. coli is regulated in response to the availability of a substrate β-glucoside by BglG induced antitermination. The RNA binding by BglG is regulated by BglF-mediated phosphorylation: in the absence of inducer, BglF phosphorylates BglG and inhibits its RNA binding activity; when β-glucosides are available, BglF dephosphorylates BglG, which now binds to its target and prevents transcription termination.

Since there have been several extensive reviews of mRNA decay within the last several years, this chapter focuses on issues that have not been completely resolved. These include the importance of RNA structural elements in mRNA decay, the existence and function of multiprotein mRNA decay complexes, the role of polyadenylation in mRNA decay, the regulation of mRNA decay, the location of mRNA decay within the cell, whether Escherichia coli is a suitable paradigm for mRNA processing and decay, the interrelationship between mRNA processing and decay, and whether all the proteins involved in mRNA decay and processing have been identified. Within these multiprotein complexes are a variety of 3' -> 5' exonucleases that are homologous to E. coli RNase PH, RNase R, and RNase D. While these enzymes in E. coli seem to be exclusively involved in the processing of tRNAs, it would not be unreasonable to think that some type of bacterial exosome might exist to promote 5′ → 3′ mRNA decay. Oligoribonuclease is responsible for degrading the very short oligoribonucleotides that are no longer substrates for PNPase, RNase II, or RNase R. With the exception of RNase E, RNase G, RNase III, and possibly yet to be identified endonucleases, all the other RNases in E. coli initiate degradation of mRNAs at the 3' terminus. mRNA decay and processing play integral roles in the regulation of bacterial gene expression.

The study of homologous recombination between plasmids, or between a plasmid and the chromosome, revealed that the RecFOR pathway is less of a poor cousin than first thought. When the exquisite sensitivity to DNA damage of the first recombination-deficient mutants was found, it became clear that homologous recombination might be the only way to repair certain DNA lesions. Generally, the stronger the defect in homologous recombination, the higher the sensitivity to DNA damage. In Escherichia coli, chromosomal lesions are repaired by homology-guided strand exchange between sister chromatids. The evidence in support of this notion comes in three forms. First, physical connections between parental and daughter strands, associated with lesion repair, can be detected. Second, repair of chromosomal lesions is not observed in recA mutants. Third, DNA damage stimulates homologous recombination although the structure of chromosomal lesions in this case is unspecified. Single-stranded DNA-binding protein (SSB) complexes single-stranded DNA (ssDNA), facilitating its subsequent use in replication and in degradation and repair pathways of DNA metabolism. Chromosomal dimerization in E. coli creates a chromosomal lesion, because it prevents segregation of the replicated chromosomes into daughter cells. The understanding of the formation of replication-dependent chromosomal lesions is still primitive. There is one in vivo study on the structure of stalled replication forks, a report documenting replication fork reversal in vivo, as well as a few reports of replication fork reversal in vitro, likely to be an artifact of DNA isolation.

This chapter focuses on the bacterial RecA proteins, which have at least three major roles. The first function involves a direct participation in the central steps of recombination, via the DNA strand exchange activity. Second, RecA protein has a role in regulation. As a regulatory function, the RecA protein facilitates the autocatalytic cleavage of the LexA repressor and certain other proteins to induce the SOS response to DNA damage. Finally, the RecA protein participates in yet another type of repair process. Late in the SOS response, especially when DNA damage levels are particularly high and nonmutagenic DNA repair is insufficient to get replication restarted, a need arises to restart replication via lesion bypass. The known biochemical activities of the RecA protein parallel these cellular roles. These include binding to DNA, ATP hydrolysis, filament formation, DNA strand exchange, and the coprotease activity. The nucleation of RecA protein on single-stranded DNA (ssDNA) is slowed considerably if the DNA is bound by the Escherichia coli single-strand DNA-binding protein SSB. The capacity to promote uniquely unidirectional DNA strand exchange reactions, to bypass significant structural barriers, and to promote four-strand exchange reactions is so far unique to the bacterial RecA proteins, and all of these processes require ATP hydrolysis.

Interaction with χ affects the helicase activity of RecBCD enzyme. Recognition of χ causes the enzyme to pause briefly at χ and to resume translocation after the χ site, but at a rate that is reduced by approximately twofold. In response to χ the RecBCD enzyme accomplishes both tasks essential for initiation of homologous recombination: (i) it recesses the double-strand break (DSB) to produce an ssDNA-tailed duplex DNA with χ at its terminus, and (ii) it catalyzes formation of the RecA nucleoprotein filament on the ssDNA produced. Interestingly, the efficiency of conjugational and transductional recombination by the RecF pathway in the recBC sbcBC cells is similar to that of the RecBCD pathway in wild-type cells, showing that the machinery of this pathway can be as productive as that of the RecBCD pathway. The loading of RecA protein is an essential aspect of recombination in the RecBCD pathway. On the other hand, recB recF double mutants are deficient in recombination between chromosomal direct repeats, suggesting that both RecBCD and RecF pathways play major roles in recombination. Homologous recombination can be initiated at either DSBs or single-strand DNA gap (SSG) in duplex DNA. Two major pathways are responsible for homologous recombination in wild-type E. coli: The RecBCD pathway is specific for the recombinational repair of DSBs, and in the wild-type cells, the RecF pathway is primarily used for recombination that initiates at SSGs.

Holliday junctions are resolved into recombinant duplex DNA species by a class of structure-specific endonucleases known as the Holliday junction-resolving enzymes. The primary cellular resolving enzyme in bacteria is RuvC, which is the main focus of this chapter. The author also talks about the RusA protein, which may act as an alternative to RuvC in some bacterial species, and attempts to place RuvC in a wider context based on our knowledge of other junction-resolving enzymes. The first cellular Holliday junction-resolving enzyme identified was RuvC from Escherichia coli. Homologous recombination is ubiquitous among cellular life forms and many prokaryotic and eukaryotic viruses, and wherever Holliday junctions are formed, junction-resolving enzymes can be confidently expected. Resolving enzymes recognize the branched structure of the Holliday junction and introduce paired phosphodiester bond cleavages in opposing strands to collapse the junction, releasing nicked duplex DNA products. The study of homologous recombination and the Holliday junction was for many years the realm of geneticists. Holliday junction migration work has largely been driven by studies of the E. coli RuvABC resolvasome, emphasizing the continuing utility of bacteria as a model system to study some of the most interesting problems in biology.

This chapter concentrates on the repair mechanism in Escherichia coli, but the lessons learned in this organism should also apply to analogous systems in other organisms. Although there are several distinct DNA mismatch repair systems, in this chapter the term is used to denote the MutSLH system. DNA polymerase III, the replicative enzyme, catalyzes resynthesis of nucleotides and ligation followed by Dam methylation to complete the process. An alternative to the futile cycling model based on double-strand DNA breaks (DSBs) recombinational repair is described in the chapter to explain how mismatch repair sensitizes E. coli dam mutants (and human cells) to methylating agents and cisplatin. In dam mutants there is constant repair of DSBs, and the recombinational capacity of the cell is probably near its maximum. This conclusion is based on the higher basal level of transcription of certain SOS genes in dam cells, suggesting that one or more of the RecA or RuvA or RuvB proteins is limiting. The hypothesis that dam bacteria are sensitive to these agents because of inability to repair all DSBs is quite plausible. An important common theme is the requirement for replication forks to stall or collapse at lesions. The hyperrecombination phenotype is explained by the increased number of DSBs leading to increased initiation of recombination. Together with the roles for Dam methylation in controlling transcription initiation and its role in regulating initiation of chromosome replication and its synchronization, almost all the phenotypic properties can now be explained.

In excision repair, damaged DNA is recognized as altered and the damage is cut out. Two types of excision repair, each with important variations, can be distinguished. The first, base excision repair (BER), uses particular enzymes, the DNA N-glycosylases to sense specific damaged bases. The second, nucleotide excision repair is a multiprotein system which recognizes generalized deformation in the DNA. The clear requirement for the RecA protein is explained by the need to support the strand exchanges required for recombination and for the filling of gaps left by the blockage of DNA synthesis along one strand only at the site of pyrimidine dimer or other lesion. The Escherichia coli genome codes for at least five different DNA polymerases. The dynamics of the bypass process involving the different DNA polymerases are described in this chapter. The complete replication system is in place and DNA is being replicated, either by progression of the complex along a chromosome, as usually thought, or, as has been suggested for Bacillus subtilis, by the chromosomal DNA moving through a fixed replication site. E. coli has a variety of enzymes which cooperate to detect and remove damage from the DNA. Operation of these mechanisms is dependent on the location of the lesions with respect to DNA growing points. Lesions far from such growing points are detected either by small DNA glycosylases that continually test the DNA for aberrant bases or by the UvrA2UvrB protein complex, which detects helix distortions and searches the distorted helix for altered bases.

Strand mispairing interactions between repeated DNA sequences provoke a host of mutations and genetic rearrangements in bacteria. Repeated DNA sequences in either direct or inverted orientation can promote misalignment-mediated mutations and rearrangements by very similar mechanisms. This chapter catalogs several types of misalignment-mediated genetic changes in bacteria, with an emphasis on the mechanisms by which they occur. Systematic study of misalignment-mediated mutation and genetic rearrangements has revealed many elements of the mechanisms of these processes, including the integral role for DNA replication. The impact of these misalignment processes on bacterial physiology and genomic evolution is also discussed in this chapter. Sister chromosome exchange (SCE)-associated misalignment has been studied only on plasmid replicons. When recA-independent deletions or expansions between tandem repeats are selected on plasmids, concomitant replicon dimerization is often seen. Single-strand annealing (SSA) contributes as a major pathway to rearrangements in eukaryotes and in bacteriophage-infected Escherichia coli but may occur efficiently only under restricted circumstances in normal bacterial growth. Short spurious repeats can be sites for deletion, duplication, or inversion; these processes create larger repeats that can lead to even higher rates of rearrangements. These rearrangements can occur independently of the bacterial homologous recombination pathways and are dependent on the length and perfection of the repeats, as well as their proximity.

This chapter talks about four different families of transposable DNA elements and describes how transposon-mediated recombination molds the organization of the bacterial chromosome. The four families were divided based on the proteins they encode for their mobility: (i) the DDE transposons, which include the majority of the classical bacterial elements such as IS3, IS50 (Tn5), IS10 (Tn10), Tn3, and phage Mu; (ii) the rolling circle transposons, called Y2-transposons and include IS91; (iii) the Y-transposons, which include the conjugative transposon Tn916; and (iv) the Stransposons, a newly recognized family that includes IS1535, IS607, and themobilizable transposon Tn4451. The mechanisms of transposition for each of these four families are very different, although the end products may often look quite similar. Their features were summarized in order to illustrate the different ways a transposon can move from one site to another and the different types of chromosomal rearrangement they can create. It is thought that immunity plays a key role in protecting a transposon from the damaging effects of its own transposition. Immunity in Tn7 parallels that of Mu, with TnsB and TnsC playing the roles of MuA and MuB. The author has compared and contrasted four families of transposable elements found in bacteria. Each family has developed distinct ways of translocating defined segments of DNA, although there are some unifying themes and many of the end products look identical.

This chapter discusses potential mechanisms for linking phage Mu transposition with cell physiology. Derepression of transposition can potentially benefit the host under conditions of stress, and these mechanisms can be part of the cellular stress response. Bacteriophage Mu is a model of regulated transposition, for it functions within its host as a fully active transposon as well as a virus. Transposition of the Mu genome into the host chromosome establishes lysogeny and replicates Mu DNA for lytic development. Upon completion of strand exchange, the transpososome remains in a tight complex with the two Mu ends in what is known as the type II transpososome or the strand-transfer complex (STC), posing as an impediment to the assembly of a replisome. The processes in DNA replication relevant to potential mechanisms in Mu derepression and those properties providing insights about Mu's relationship with its host are summarized in this chapter. Two types of repressor mutants which induce lytic development in Mu lysogens have provided insight as to how Mu derepression may be triggered. The C-terminal domain (CTD) of Rep plays an important role not only in eliciting thermolability of cts DNA binding domain (DBD) mutations present in cis but also in promoting Rep degradation induced by Vir expressed in trans. Recent evidence implicates a role for the Mu repressor CTD in S derepression and regulation of transposition. The CTD’s influence on DNA binding as well as repressor degradation represents two potential pathways by which derepression may be triggered.

This chapter presents the processes that can lead to the dimerization of replicons and discusses the mechanisms that ensure their resolution. It also discusses how chromosome dimer resolution is integrated into other aspects of DNA processing during the bacterial cell cycle. The chromosomes and linear plasmids of Borrelia are linear with hairpin ends. Bidirectional replication is initiated internally and, when complete, generates a circular dimer of the original linear replicon. In this circular dimer, the hairpin sites of the parental DNA are converted into palindromic “telomere” sites, which are used for chromosome dimer resolution. Xer site-specific recombination, which is responsible for chromosome dimer resolution, ensured their stable inheritance within Escherichia coli. In Borrelia and in bacteriophage N15 of E. coli, resolution of chromosome dimers is due to the action of a single enzyme, ResT or TelN, respectively. During tyrosine recombinase-mediated site-specific recombination, two tyrosine recombinase molecules bind cooperatively to ~30-bp specific core recombination sites in the DNA. The XerC and XerD site-specific recombinases function in chromosome dimer resolution by adding a single crossover at dif, a specific 28-bp core site located in the region of termination of replication of the chromosome. In vivo and in vitro studies show that in the absence of FtsK, Holliday junctions (HJs) are created and resolved back to the original substrate in cycles of XerC-mediated strand exchanges. The realization that Xer recombination uses different strategies to ensure resolution selectivity during plasmid and chromosome dimer resolution demonstrates the sophistication that has developed during bacterial evolution.

This chapter focuses on the structure and associated functional features of the linear bacterial chromosomes of Borrelia species and Streptomyces species. Telomere resolvases have also been referred to as protelomerases, for prokaryotic or proteic telomerases. Based upon the similarity of the telomeres on the chromosome to those on the linear plasmids, replication of the linear plasmids is also expected to be bidirectional from an internal origin. The recombination model suggests that homologous recombination would be essential for chromosomal replication and thus viability. Nevertheless, it is noteworthy that the recombination model in theory is blind to the sophisticated secondary structures at the 3’ overhangs during DNA replication. The S. coelicolor chromosome sequence, however, contains an open reading frame encoding another α subunit of PolIII. There are also two copies of β and γ/τ subunits and three of ε subunits. The meaning of this multiplicity is not clear, but it is unlikely that an extra PolIII enzyme functions in terminal patching, because the DNA strand synthesized in terminal patching is expected to be relatively short , not necessitating the high processivity of such enzyme. The linear replicons of Borrelia and Streptomyces, despite their identical topology, appear to be highly diversified in their structures and modes of replication. These two organisms use varied but highly effective mechanisms to ensure that the ends of their linear DNA molecules are faithfully replicated, thereby eliminating the need for circularity of their DNA.