Homologous recombination promotes the pairing between identical — or nearly identical – DNA sequences and the subsequent exchange of genetic material between them. It is an important and widely conserved function in living organisms, from bacteria to humans, that serves to repair double-stranded breaks or single-stranded gaps in the DNA, arising as a consequence of ionizing radiations, ultraviolet (UV) light, or chemical treatments creating replication-blocking adducts (Kuzminov, 1999). More recently, homologous recombination functions were also found in bacteria to rescue replication forks that have stalled for various reasons, such as a missing factor (e.g., the helicase), or a particular difficulty upstream of the fork, such as supercoiling or intense traffic of proteins (Michel et al., 2001).

Besides its molecular role, homologous recombination has played a major role in genome dynamics, by changing gene copy numbers through deletions, duplications, and amplifications: Intrachromosomal recombination between ribosomal operons or between mobile elements scattered into the genome leads to deletion or tandem duplications of large regions within the genome, up to several hundred kilobases (Roth et al., 1996). The duplications are unstable. Mostly they recombine back to the parental organization, and, therefore, remain undetected, except when appropriate selection, by gene dosage mostly, is exerted (Petes and Hill, 1988). In contrast, such duplications are ideal substrate for the diversification of genes: One gene is kept intact whereas the other is mutagenized, which leads to the birth of gene families.

Until the late 1970s, it was believed that the majority of gene transfer events in bacteria were mediated by plasmids. However, at that time, Don Clewell and co-workers isolated a strain of Enterococcus faecalis that could transfer tetracycline resistance in the absence of plasmid DNA (Tomich et al., 1979). The element responsible was an 18-kb segment of DNA that was integrated into the bacterial chromosome. As well as being capable of conjugative transfer to a new host, this element was also capable of intercellular transposition, so the term conjugative transposon was coined and this particular element was designated Tn916 (Franke and Clewell, 1981). Subsequently, it has become apparent that conjugative transposons are probably ubiquitous and are highly heterogeneous. This heterogeneity in form and function has led to some confusion and controversy about what a conjugative transposon actually is and how these elements should be named. This issue has more recently been addressed in a number of review articles (Burrus et al., 2002; Mullany et al., 2002; Osborn and Boltner, 2002). Therefore, for the purposes of this chapter, we define conjugative transposons in the loosest possible terms, as discrete DNA elements, usually integrated into the bacterial genome, which can transfer from a donor to a recipient cell by conjugation.

Because conjugative transposons have such a broad host range, they are very important in bacterial evolution (i.e., in disseminating genes between distantly related organisms, induction of deletion, rearrangements) and as a substrate for recombination events (Beaber et al., 2002; Mahairas and Minion, 1989; O'Keefe et al., 1999; Swartley et al., 1993).

Competence for genetic transformation is a physiological state that enables the uptake of exogenous DNA. Although competence is widespread in nature (Lorenz and Wackernagel, 1994), it appears to be a variable phenotype because natural isolates of a given species may or may not be transformable, and genome sequencing has revealed the presence of competence genes in isolates that are not known to be transformable.

WHAT USE IS COMPETENCE?

In this chapter, we consider three disputed hypotheses that have been advanced for the biological role of competence: DNA uptake for new genetic information, DNA uptake for repair, and DNA uptake for nutrition.

DNA for genetic diversity

The evolutionary pressure for the maintenance of competence genes may be explained by the advantages gained from the acquisition of fitness-enhancing genes; competence may expand the repertoire of genes available to improve the chances of survival in harsh conditions. Several examples of the acquisition by transformation of new genes that confer a selective advantage have been suggested. For instance, in Neisseria gonorrhoeae, in which transformation is the only known mode of DNA transfer, the expression of new allelic variants may facilitate antigenic variation, which presumably plays a role in the evasion of the host immune response. Thus, although new pilin variants may be formed by intracellular recombination between a silent gene segment and an expressed pilin gene, the intercellular exchange of pilin alleles may also occur by transformation (Gibbs et al., 1989; Seifert et al., 1988).

The mechanisms of homologous recombination in bacteria (described in Chapter 1) are ancient and highly conserved. The basic requirement is two DNA sequences that share sequence identity over a minimum distance, typically at least 40 to 50 nucleotides (Shen and Huang, 1986; Watt et al., 1985). These identical sequences are brought together to create, and eventually resolve, a recombinant molecule, by the actions of enzymes such as RecA, RecBCD/RecF, and RuvABC, or their equivalents (Kowalczykowski et al., 1994). Homologous recombination serves several purposes in the cell. The most fundamental of these, according to current understanding, is to facilitate the completion of chromosome replication (Cox, 2001; Smith, 2001). Replication forks frequently stall or break, and homologous recombination can solve this problem using the sister homolog as a template (Kuzminov, 1995). A related problem is that after replication bacterial chromosomes are frequently entangled and require homologous recombination to disentangle them.

In addition to its housekeeping roles, homologous recombination can shuffle the order of genes in a genome by recombination between repetitive sequences. It can also facilitate the incorporation of foreign DNA, although this is also achieved by site-specific recombination (Ochman et al., 2000). Lateral DNA transfer in bacteria can alleviate the effects of Muller's ratchet (Andersson and Hughes, 1996; Muller, 1964), and provide bacteria with access to a very large gene pool potentially containing important innovative properties (Ochman et al., 2000).

Molecular biologists have long used viruses, plasmids, transposons, and other “vectors” as tools to directly manipulate the genetic makeup of experimental organisms. In nature, these tool vectors originated in species, usually bacteria, as facilitators of horizontal (also known as lateral) gene transfer (HGT). In contrast to vertical inheritance, where the transmission of genetic material occurs vertically from parent to offspring, HGT refers to the horizontal exchange of genes between distantly related strains and species. As described in this volume, there are many examples of HGT between species of bacteria, such as that mediated by plasmids and phages, which bear genes responsible for pathogenicity and antibiotic resistance. HGT is also known to occur in eukaryotes; for example, DNA transposons have been suggested as being horizontally transferred between different species of the fruitfly Drosophila (Bushman, 2002). These are examples of HGT on a relatively recent evolutionary timescale. However, HGT might have had a pivotal evolutionary role in more ancient times. Comparative analyses of molecular data that are exploding from genome sequencing projects indicates that HGT might have been the main driving force behind the evolution of cellular life (Brown, 2003).

The reason for believing the occurrence of ancient HGT is relatively simple. In an evolutionary context, genes are not found where they are expected to be. The most fundamental subdivisions of living organisms are the three urkingdoms or domains of life: the Archea (traditionally called “archaebacteria”), Bacteria (traditionally called “eubacteria”), and Eucarya (interchangeable here and elsewhere with the term “eukaryote”; Woese, Kandler, and Wheelis, 1990).

Infections caused by microbial pathogens are a major global health problem not only in developing countries, but also in the industrial world. Similarly, there are numerous diseases of animals and plants due to bacterial infections. Pathogenic bacteria can be found among various bacterial species, and the determination of the factors that are responsible for virulence of a certain bacterium has long been one of the main interests in microbial research. In the early 1990s, new insights into the evolution of bacterial pathogens were gained by the development of the concept of pathogenicity islands (PAIs), which are the topic of this chapter. Since then, the advances in genomics have led into a new area of pathogen research, with increasing knowledge of completely sequenced bacterial genomes.

The prokaryotic genome can be generally divided into a core gene pool encompassing those genes that encode essential functions, such as DNA replication, cell division, nucleotide turnover, and key metabolic pathways, and a flexible gene pool containing genes that are only required under certain environmental conditions. The core genes are normally encoded in stable regions of the chromosome and exhibit a relatively homogeneous G+C content. In contrast, genes of the flexible gene pool are often found on mobile genetic elements and are preferentially transmitted between different organisms by natural transformation (the uptake of naked DNA), phage-mediated transduction, or conjugation (the unidirectional transfer of DNA from a donor to a recipient via intimate cell-to-cell contact).

Bacteria owe their ability to thrive in diverse and ever-changing conditions to their extraordinary faculty of adaptation. They have developed many strategies to adjust to new environments. These mechanisms include random modifications within their genome, such as point mutation, duplication, deletion, insertion, and acquisition of new DNA (e.g., lateral gene transfer). These multiple events generate a heterogeneous microbial population containing numerous novel phenotypes. Whenever the environment changes, a subpopulation more apt to survive in these new conditions emerges, thus allowing bacteria to thrive, for example, by acquiring resistance to antibiotics. However, as the intensity, duration, and nature of stress are extremely variable, the optimal response to new environmental conditions may be unpredictable. The means by which bacteria either respond to stress, such as exposure to toxic/inhibitory compounds or starvation, or avoid detection by the host's immune system are crucial for their survival. The spontaneous mutation rate is usually insufficient for allowing an efficient adaptation to these changes. However, certain bacterial populations contain hypermutator strains exhibiting highly increased rates of spontaneous mutations, therefore promoting adaptation to changing environments (Taddei et al., 1997). Nevertheless, this benefit may disappear once adaptation is achieved because the evolved genotype may have accumulated irreversible mutations that are detrimental in other conditions (Giraud et al., 2001a, 2001b).

Alternatively, bacteria have developed adaptation strategies based on DNA rearrangement events restricted to specific genomic regions. These defined loci allow bacteria to generate an array of phenotypic variants, whereas minimizing detrimental effects of random mutations on fitness.

Recombination provides a means for creating genetic variety. Exchange of information within gene pools expands their diversity and enhances the choices available for natural selection to act on. Recombination can be broadly divided into two classes: the highly pervasive “homologous” recombination and the more specialized “site-specific recombination.”

HOMOLOGOUS RECOMBINATION

Before we discuss site-specific recombination, a brief overview of general (or homologous) recombination is useful for an appreciation of the distinctions between the two systems. Homologous recombination is a nearly universal mechanism employed by living organisms to reshuffle their genetic information. Within a cell, recombination can occur between two homologous chromosomes, between two sister chromatids formed by DNA replication, and between extrachromosomal elements such as plasmids or viral genomes. In eukaryotes, the rate of recombination during mitosis is relatively low and is markedly increased during meiosis. In fact, genetic exchange between homologs and chiasma formation appear to be a prerequisite for the proper reductional segregation of chromosomes and the generation of haploid gametes. The consequences of the resulting genetic configuration and the corresponding fitness contribution to an individual will be manifested directly, and almost immediately, in a haploid organism. For a diploid organism, the expression of the novel genetic makeup must await the fusion between the male and female gametes to produce a zygote. Recombination is therefore one of the forces that drive Darwinian evolution.

It is a curious fact that two unrelated families of enzymes have evolved that promote conservative site-specific recombination. Elsewhere in this volume (see Chapter 2), a large family of “tyrosine recombinases” is described, of which phage lambda integrase is the most famous and senior member. The subject of this chapter is a second large family, the “serine recombinases.” The names come from the conserved residue of the recombinase that provides the nucleophile to attack and break the DNA phosphodiester backbone (see “Tn3 and γδ resolvases: cointegrate resolution” section). Although the serine and tyrosine recombinases are unrelated in sequence, structure, or mechanism, there is no obvious distinction between their biological functions. Why two very different types of site-specific recombinases have survived eons of natural selection, yet continue to play similar roles, remains a mystery. Serine recombinases are widespread in the Eubacteria and Archea, but not in Eukarya, where the few examples found so far may be of recent bacterial origin.

A serine recombinase can be identified by similarity of parts or all of its primary amino acid sequence to that of one of the archetypal members of the family [e.g. Tn3 resolvase (Fig. 3.7)]. Several hundreds of such proteins can now be predicted from available DNA sequences (reviewed by Smith and Thorpe, 2002). The relations of the members of the family are discussed later in this chapter.

Introns and retroelements are hallmarks of eukaryotic genomes, but they are also found in bacteria. Here the different types of bacterial introns and retroelements are summarized, including group I introns, group II introns, archaeal bulge-helix-bulge (BHB) introns, intervening sequences (IVSs) in rRNAs, retrons, and diversity generating elements (DGRs). Except for the retroelements, these elements are evolutionarily unrelated, but nevertheless share intriguing properties. The elements all appear mobile within and among bacterial genomes, and in general, do not have clear phenotypic consequence to their host cells. It is possible that introns and retroelements spread from bacteria to eukaryotes as selfish DNAs or were present in the common ancestor of bacteria and eukaryotes.

INTRODUCTION

Introns and retroelements are typically considered components of eukaryotic genomes, because they were discovered in eukaryotes and are particularly abundant in higher eukaryotes. The human genome, for example, contains roughly 200,000 introns and nearly 3 million retroelements (including SINEs), dwarfing the number of functional genes, which are estimated at 30,000 (International Human Genome Sequencing Consortium, 2001). Together, introns and retroelements make up nearly half of the human genome and constitute the major types of “junk DNAs.”

In bacteria, the major types of “junk DNAs” are transposons and prophages, which can constitute anywhere from <1% to 7% of a genome (e.g., Glaser et al., 2001). Introns and retroelements are comparatively rare in bacteria, but they have generated substantial interest because of their parallels to eukaryotic introns and retroelements.

The F factor is often associated with Escherichia coli and appears to have been adapted by the bacterial host to act as an agent of genetic exchange and evolution. F encodes a type IV secretion system (T4SS) that enables bacterial conjugation, the transfer of DNA from a donor F+ to a recipient F− cell. The delivery of DNA containing either host or foreign genes has important consequences for the bacterium, allowing it to enlarge or modify its genetic content and rapidly adapt to an environmental niche. Unlike other plasmids, the conjugative functions of F and F-like plasmids appear to be controlled by a complex regulatory network that involves many host proteins resulting in a symbiotic relation between F and its host. This chapter outlines the predicted and known functions for all the genes on the F plasmid and its close relatives, and describes our current knowledge about the regulation of F conjugation.

BRIEF HISTORY OF THE F PLASMID

The discovery of horizontal gene transfer between bacteria can be attributed to the work of Lederberg and Tatum (1946), who observed that different strains of E. coli K-12 could be genetically and phenotypically altered when mixed together. A series of experiments led to the conclusion that direct contact between bacteria was required in order for genetic material to be transferred between the cells (Davis, 1950). This transfer was found to occur in one direction, from donor to recipient cells, by a mechanism contained within the donor cells (Hayes, 1952).