Are involved in the transfer of genetic information from one bacterial cell to another?

Conjugation is a type of bacterial ‘mating’ in which DNA is transferred from one bacterium to another

Conjugation is dependent upon thetra genes found in ‘conjugative’ plasmids, which, among other things, encode instructions for the bacterial cell to produce a sex pilus – a tube-like appendage which allows cell-to-cell contact to ensure the protected transfer of a plasmid DNA copy from a donor cell to a recipient (seeFig. 2.18). Since thetra genes take up genetic space, ‘conjugative’ plasmids are generally larger than non-conjugative ones.

Occasionally, conjugative plasmids such as the fertility plasmid (F plasmid or F factor) ofE. coli integrate into the bacterial genome (e.g. facilitated by identical IS elements on both molecules as noted earlier), and such integrated plasmids are called episomes. When an integrated F episome attempts conjugative transfer, the duplication-transfer process eventually moves into regions of adjacent genomic DNA, which are carried along from the donor cell into the recipient. Such strains, in contrast to cells containing the unintegrated F plasmid, mediate high-frequency transfer and recombination of genomic DNA (Hfr strains). However, conjugation with Hfr donor cells does not result in complete transfer of the integrated plasmid. Thus, the recipient cell does not become Hfr and is incapable of serving as a conjugation donor. The circular nature of the bacterial genome and the relative ‘map’ positions of different genes were established using interrupted mating of Hfr strains.

When a non-conjugative plasmid is present in the same cell as a conjugative plasmid, they are sometimes transferred together into the recipient cell by a process known as mobilization. Conjugative transfer of plasmids with resistance genes has been an important cause of the spread of resistance to commonly used antibiotics within and between many bacterial species, since no recombination is required for expression in the recipient. Of all the mechanisms for gene transfer, this rapid and highly efficient movement of genetic information through bacterial populations is clearly of the highest clinical relevance.

Bacterial Conjugation

Bacterial conjugation is one of the three major known modes of genetic exchange between bacteria, the other two being transduction and bacterial transformation. Of these three modes, conjugation is the only one that involves cell-to-cell contact. J. Lederberg and E. L. Tatum first reported such transfer in 1946 in Escherichia coli. The discovery resulted from a deliberate search for sexual recombination in bacteria, in which progeny carry genetic markers from two parents. Bacterial conjugation is a sexual mode of genetic transfer in the sense that chromosomal material from two sexually distinct cell types is brought together in a defined and programmed process. However, in bacterial conjugation, the process involves only a portion (usually small) of the genome of one of the cells (the donor) and the complete genome of its sexual partner (the recipient), as opposed to sexual union in most higher organisms, which involves an interaction between the entire set of chromosomes from both of the parental cells. Thus, genetic transfer in bacterial conjugation is partial, and it is polar in most cases, wherein genetic material moves unidirectionally from the donor cell into the recipient cell followed by separation of the cells and further changes in the organization or recombination of the combined genetic material within the recipient cell. The transfer of genetic material can take several minutes or more (up to several hours).

Pili

Most of the Enterobacteriaceae produce thin nanofilaments extending from the bacterial surface called pili—also known as fimbriae—that mediate autoaggregation and adhesion to host cells.28 Furthermore, pili facilitate bacterial conjugation and are often encoded by plasmids—capable of harboring virulence and antimicrobial resistance genes—to mediate intercellular contact and genetic exchange.29 The family Enterobacteriaceae features a variety of pilus types, which differ in morphology and function. An individual strain may produce multiple different pili even of the same type.30–32

Chaperone-usher type pili are common among the Enterobacteriaceae and include the ubiquitous type 1 pili (seeFig. 218.2B). Assembly begins with secretion machinery exporting subunits of the major structural protein pilin to the periplasmic space to bind with chaperones to prevent premature subunit interactions. The pilin-chaperone complexes are then delivered across the outer membrane—in a specific order starting with the tip—by a membrane channel protein known as the usher.33,34 Most chaperone-usher type pili feature a rigid rod composed of a helical array of pilin joined end to end with a thinner, more flexible tip.35,36 The tip often features adhesin proteins that serve as critical virulence factors.

Type IV pili are also widespread among the Enterobacteriaceae, often forming ropelike bundles expressed at the poles of the organism37–41 (seeFig. 218.2C). The type IV pilin protein is processed to its mature form by a dedicated prepilin peptidase that alsoN-methylates the amino-terminal residue.42 Type IV pili are retractable, which facilitates aggregation and disaggregation and a type of locomotion called “twitching motility.”43

Introduction

Bacterial conjugation was first described by Lederberg and Tatum in 1946 as a phenomenon involving the exchange of markers between closely related strains of Escherichia coli. The agent responsible for this process was later found to be a site on the chromosome called the F (‘fertility’) factor. This finding was the basis of bacterial genetics in the 1940s and 1950s and was used extensively in mapping the E. coli chromosome, making it the preeminent prokaryotic organism at that time. It was also shown that F could excise out of the chromosome and exist as an extrachromosomal element or plasmid. It was capable of self-transfer to other bacteria and could cotransfer the chromosome, a serendipitous function of F, and integrate randomly into its host’s DNA. The F sex factor of E. coli also imparted sensitivity to bacteriophages that required the F pilus, which is encoded by the F transfer region, as an attachment site during infection. In the 1960s a number of other conjugative plasmids were isolated, many carrying multiple antibiotic resistance markers. These plasmids were termed R (‘resistance’) factors and were found in many instances to repress pilus expression and conjugation by F, a process termed fertility inhibition (fi+). The number of conjugative plasmids discovered has grown tremendously in the last few decades and includes self-transmissible plasmids isolated from Gram-negative and -positive bacteria as well as mobilizable plasmids. Conjugative transposons or integrating conjugative elements (ICEs), which move between cells using a conjugative mechanism, excise and integrate into the host chromosome via a process reminiscent of lysogenic phages; an example of a conjugative phage has been described for Staphylococcus aureus.

In general, the transfer and replication functions of these mobile elements are often physically linked and the type of transfer system is closely aligned with the nature of the replicon that is described by incompatibility groups (Inc). An excellent summary of the properties of many conjugative plasmids is given in Shapiro (1977).

Bacterial conjugation is now realized to be one of the principal conduits for horizontal gene transfer (HGT) among microorganisms. The process is extremely widespread and can occur intra- and intergenerically as well as between kingdoms (bacteria to yeast or to plants). DNA sequence analysis has revealed that conjugation, and in some cases transformation, two of the main conduits for HGT, are effected by a transenvelope protein complex that belongs to the type IV secretion system (T4SS). The effect of this process on evolution has been immense with bacteria rapidly acquiring traits both good (hydrocarbon utilization) and bad (antibiotic resistance, toxins). Once again, bacterial conjugation is at the forefront of microbiology but this time the emphasis is on the process itself rather than its utility as a geneticist’s tool. Excellent reviews of the topic are provided in The Horizontal Gene Pool, Bacterial Plasmids and Gene Spread (C.M. Thomas, ed.) and Plasmid Biology (Phillips, G. and Funnell, B., eds.).

Transferable Resistance

Transferable resistance usually results from bacterial conjugation and the transfer ofplasmids (extrachromosomal DNA) that confer drug resistance (seeFig. 37.4B). Transferable resistance, however, can also be mediated by transformation (uptake of naked DNA) or transduction (transfer of bacterial DNA by a bacteriophage). Bacterial conjugation enables a bacterium to donate a plasmid containing genes that encode proteins responsible for resistance to an antibiotic. These genes are calledresistance factors. The resistance factors can be transferred both within a particular species and between different species, so they often confermultidrug resistance. The various species need not all be present during the period in which the antibiotic is administered. Studies have shown that resident microflora of the human body can serve as reservoirs for resistance genes, allowing the transfer of these genes to organisms that later invade and colonize the host.

Several genes responsible for drug resistance have been cloned, and the factors that control their expression are being studied. In the future, drugs that block the expression of these genes may find use as adjunct therapy for infectious diseases. For example, it may be possible to develop antisense nucleotides that block the transcription or translation of genes that encode proteins responsible for drug resistance.

Historical Aspects

The discovery of bacterial conjugation in 1946 was hailed by Salvador Luria in 1947 as “probably among the most fundamental advances in the whole history of bacteriological science,” even before the most basic facets of the process were known. This proved prophetic and even an understatement since studies using conjugation led to seminal new concepts such as the existence of circular bacterial chromosomes, transferable antibiotic resistance, the chromosomal location and conjugational inducibility of lysogenic bacteriophages, multiple pathways of genetic recombination (including plasmid–chromosome recombination), and fundamental aspects of the regulation of operons as deduced from experiments with Hfr crosses and merodiploids involving the lac, trp, and other operons. For their work on conjugation and related topics, Lederberg, Tatum, and Beadle received the Nobel Prize in 1958.

Bacterial Conjugation

Transfer of genetic material occurs during the process of bacterial conjugation. During this process, DNA plasmid is transferred from one bacterium (the donor) of a mating pair into another (the recipient) via a pilus. The transfer can take about 90–150 minutes, and can be observed directly by video-enhanced microscopy [45] (Fig. 2.23). As one bacterium of a mating pair approaches, pilus binding takes place (Fig. 2.23a), followed by pilus retraction and stabilization of wall-to-wall contact (Fig. 2.23b). DNA transfer follows (Fig. 2.23c) and then the cells separate (Fig. 2.23d). During wall-to-wall contact of the mating bacteria, DNA transfer takes place. In order to observe the delicate structure of the contact point by electron microscopy, Dürrenberger et al. [45] used rapid freezing and cryosubstitution. Serial sections through the point of contact, termed the conjugational junction, did not reveal specific substructures, such as plasma bridges with fusion of membranes or cell wall to mediate DNA transfer (Fig. 2.24). The conjugational junction resembled the morphology of tight junctions in eukaryotic cells. It is likely, therefore, that the pilus initially triggers a signal for the mechanism of DNA transfer to be initiated, possibly by the formation of a continuous protein channel across the periplasmic gel of the recipient cell wall [45]. Subsequently, DNA is transferred to the recipient cell from the donor through the channel.

Plasmids and Horizontal Gene Transfer

One of the primary mechanisms for HGT is gene acquisition via the transfer of plasmids during bacterial conjugation. The term plasmid was first introduced by Lederberg (1952) and defined “as a generic term for any extrachromosomal hereditary determinant.” However, more modern definitions of plasmids include the caveat that these extrachromosomal DNA determinants are self-replicating genetic elements. Several plasmids are self-transmissible in that they encode the necessary machinery for transferring a copy of their DNA to adjacent cells by means of conjugation. The genes carried by plasmids are typically not essential for normal cellular function but, as in the case of antibiotic resistance, could be vital to survival in certain environments. Most plasmids are circular dsDNA but linear plasmids also exist (Stewart et al., 2005; Hinnebusch and Tilly, 1993); they typically consist of (1) a ‘backbone,’ which contains the genes necessary for self-replication, maintenance, control, and conjugative transfer and (2) various ‘accessory’ genes that provide other functions to the host bacterium. It is these plasmid-borne accessory genes that contribute, at least in part, to the rapid spread and emergence of traits across Archaea and Bacteria.

Numerous types/classes of traits are encoded for by the accessory genes of plasmids. As previously mentioned, antibiotic resistance is one of the most concerning of the traits. The transfer of antibiotic resistance by means of plasmids was first documented between members of the Enterobacteriaceae in the late 1950s and early 1960s (Leclercq, 2002; Watanabe and Fukasawa, 1960). However, as shown in Figure 1, increased resistance to many drugs, and in particular ‘last line’ drugs like vancomycin (Chang et al., 2003; Weigel et al., 2003), has been spreading rapidly in the last several decades and is also tied to plasmid encoded genes (Friães et al., 2014). Resistance to toxic heavy metals (such as cadmium, cobalt, silver, lead, and mercury) is also often encoded by plasmids. For example, the plasmid pWR501 has been demonstrated to transfer genes for resistance to mercury in Shigella (Venkatesan et al., 2001). Heavy metal resistance genes allow bacteria to exist and thrive in harsh, contaminated environments. Moreover, due to co-localization of metal and antibiotic resistance genes on the same plasmids, the presence of heavy metals may co-select antibiotic resistance, as suggested by recent alarming examples such as in the MRSA strain ST398 (Gómez-Sanz et al., 2013). Plasmids also often transfer virulence factors between bacteria. A classic example of virulence encoded by plasmids occurs in the Bacillus cereus group of which the species B. cereus, Bacillus thuringiensis, and Bacillus anthracis are all highly genetically related members. However, B. anthracis is the etiological agent of the dreaded anthrax disease. The plasmids pXO1 and pXO2 carry the genes that encode for the exotoxin and acid capsule (both virulence factors) of B. anthracis, respectively (Kolstø et al., 2009). Degradation of organic compounds is another function plasmids may provide bacteria. These catabolic plasmids are usually large (>50 kb) because degradation of organic compounds often requires numerous genes that are part of catabolic cascading pathways. Classical examples of this are naphthalene and toluene/xylene degradation by the plasmids NAH7 and pWWO in Pseudomonas putida, and the many catabolic plasmids of the broad-host-range plasmid group IncP-1 (Fernández et al., 2012; Top and Springael, 2003). It should be noted that this is not an exhaustive list of the many phenotypes plasmids can encode in bacteria (Top et al., 2000), just those that are often considered highly important.

Plasmids are typically classified in incompatibility groups. Incompatibility is defined as the inability of two plasmids to be vertically co-transmitted within a cell lineage for multiple generations (Couturier et al., 1988). Traditionally, this was determined using selection for plasmids and then releasing the selection pressure and observing which plasmids remained in a lineage. More recently incompatibility types are being determined by DNA sequence information on the replication region of the plasmid (Sota and Top, 2008; Carattoli et al., 2005). Plasmids of Gram-negative bacteria are typically classified into an alphabetical grouping (i.e., IncA through IncZ). Classification of plasmids from Gram-positive bacteria follows various methods, for example, the plasmids of S. aureus are broken into 15 families (Inc1 through Inc15). Needless to say, the classification of plasmids is confusing and overlap between classification systems exists (Couturier et al., 1988), but the concept of plasmid coexistence within a bacterium is important to how groups of plasmids can shape evolution in bacteria.

In broader biological terms, the relationship between plasmids and bacteria can be thought of as an interaction between species where bacteria fulfill the role of the host. Thus a great deal of research has focused on the host range of plasmids, with the delineation of having either broad or narrow host range (BHR vs NHR). The host range of a plasmid is normally defined by the group of hosts in which a plasmid can successfully replicate. This does not necessarily indicate which bacteria a plasmid may be found in. Without selective pressure, replicating plasmids may be rapidly lost from some lineages but can also be efficiently retained in other strains, even within the same species (De Gelder et al., 2007). Conjugative transfer of some plasmids can occur to bacteria or even eukaryotes in which the plasmid cannot replicate, resulting in a transfer range that is much wider than the replication range (Thomas and Smith, 1987). In terms of the evolution of unwanted phenotypes, like antibiotic resistance, BHR plasmids are the most problematic because of their ability to pass genes for such traits to numerous species of bacteria and may be the most important means of HGT between distantly related bacterial hosts (Mazodier and Davies, 1991). Thus, one of the central questions related to plasmid-mediated bacterial evolution is how host ranges and the stability of a plasmid within a host evolve.

C The bacteriophage P22 recombination system

Experiments by Norton Zinder in the Lederberg laboratory in early 1950s were designed to discover if Salmonella typhimurium could promote bacterial conjugation, as demonstrated previously in E. coli (Zinder and Lederberg, 1952). The mixing of different autotrophic mutants in genes involved in amino acid metabolism, followed by plating for prototrophs, led to the identification of one particular strain (#22 in the collection) that generated wild-type phenotypes at high frequency. Unexpectantly, it led to the discovery of a bacteriophage (now called P22) that could carry out cell-to-cell transfer of a gene, a process known today as viral transduction. P22 is a temperate bacteriophage of S. typhimurium that possesses a circularly permuted and terminally redundant chromosome (Levine, 1972; Poteete, 1988; Susskind and Botstein, 1978). One of the key steps in the life cycle of phage P22 is circularization of the infecting linear chromosome, which is promoted by the phage recombination system acting on the terminally repetitious ends of the phage DNA (Weaver and Levine, 1977). This mechanism of circularization is different from that seen with phage λ, where circularization proceeds by annealing of overlapping single-stranded cos (cohesive ends) sites. This distinction makes recombination an essential feature of P22 biology, which is not the case for phage λ. Similar to λ though, phage P22 generates concatemeric DNA for packaging (Botstein, 1968). Like λ, phage P22 generates chromosome multimers in two ways: a phage-promoted recombination event between two monomeric circles and the rolling-circle mode of DNA replication (Gilbert and Dressler, 1968; Weaver and Levine, 1977).

The recombination system of P22 consists of four proteins: Erf, Abc1, Abc2, and Arf. The ssDNA-annealing function is provided by the Erf protein (essential recombination function) (Botstein and Matz, 1970; Poteete, 1982; Weaver and Levine, 1977). The Erf protein was shown to bind to ssDNA (Poteete and Fenton, 1983). Given its ability to be replaced by the λ Beta protein in P22 phage crosses, Erf most likely possesses an ssDNA-annealing function similar to Beta. Erf was the first phage recombinase to show the ring-like quaternary structure under electron micrographs (Poteete et al., 1983). As described earlier, Erf showed projections emanating from its rings, which were identified as the C-terminal domains of the individual subunits (Murphy et al., 1987a). Erf fragments missing the C-terminal domain could still form rings, showing that this region of the protein was not required for either formation or stability of the rings (Murphy et al., 1987a; Poteete et al., 1983). However, the C-terminal domain is important for recombinase function in vivo, as P22 erf amber mutants expressing only the N-terminal ring-forming domain were deficient for growth in a recA host strain. Further analysis of purified Erf protein fragments revealed that a region important for binding ssDNA was located between the N-terminal ring-forming domain and the C-terminal domain of unknown function (Murphy et al., 1987a).

Unlike the λ Beta protein, Erf does not have an associated 5′-3′ exonuclease partner, nor does it have an accompanying protein such as λ Gam, which binds to and inhibits the RecBCD enzyme. Amazingly though, these two functions are provided by the Abc2 (anti-recBCD) protein that binds to the host RecBCD enzyme and modifies its exonuclease function so that it works cooperatively with Erf to promote recombination in vivo (Murphy, 1994; Murphy and Lewis, 1993; Murphy et al., 1987b). Effectively, the RecBCD protein is hijacked by Abc2 and made to work as part of the P22 recombination system. How this is accomplished is not clear. It is known that Abc2 binds to the RecC subunit of the RecBCD complex. In assays designed to measure in vitro nicking of RecBCD at its hot spot sequence Chi (3′GCTGGTGG5′), the Abc2–RecBCD complex was incapable of recognizing or responding to Chi (Murphy, 2000). (This inhibition of Chi activity is not relevant to P22 recombination, as P22-promoted recombination is independent of Chi.) More importantly, Abc2-modifed RecBCD still possesses dsDNA exonuclease activity, although it is qualitatively modified (Murphy, 2000). Digestion of the 5′ strand is upregulated by Abc2, not unlike a native RecBCD species after its encounter with Chi (Dillingham and Kowalczykowski, 2008). Although not yet tested, one possibility is that Abc2 prevents loading of RecA at Chi sites, thus explaining how the expression of Abc2 alone inhibits host RecABCD-promoted conjugational recombination events. It may be that Abc2 interferes with the interaction between RecA and RecBCD and, at the same time, allows the modified enzyme to cooperate with the P22 Erf protein to promote recombination, and likely other SSAPs as well. This latter supposition comes from the observation that coexpression of Abc2 with P22 Erf, λ Beta, and even the Pseudomonas phage D3 SSAP Orf-52 promotes the growth of λ red gam mutants in E. coli recA hosts (K. Murphy, unpublished observations). This hypothesis explains why Beta can substitute for Erf in a P22 infection, although Erf cannot substitute for Beta in a λ infection (Poteete and Fenton, 1984). Such a complementation would require an interaction between P22 Erf and λ Exo. It is known that the interacting partners of the λ Red and the rac prophage RecET system cannot substitute for one another (i.e., λ Beta cannot work with RecE and λ Exo cannot work with RecT to promote recombination) (Muyrers et al., 2000). If exogenous phage SSAPs interact directly with Abc2-modified RecBCD, it would suggest that the modified enzyme interacts with a motif common to a variety of phage SSAPs. Alternatively, no direct interaction between an Abc2-modified RecBCD and an accompanying SSAP is required.

Other components of the P22 system play a less defined role, but are required for full growth and recombination of a P22 phage deleted of its recombination region in Salmonella recA hosts. The Abc1 protein was defined as assisting Abc2 for the full restoration of growth and recombination in P22 mutants deficient for recombination (Murphy and Lewis, 1993; Murphy et al., 1987b). However, the ability of Abc1 to interact with RecBCD has not been studied. The Arf (accessory recombination function) protein is encoded by a small open reading frame upstream of the erf gene. It was originally not expected to be involved in recombination due to its small size (47 amino acids), the unusually acidic nature of the predicted protein (pI of 3.5), and the suspicion that the modest effects of a deletion identifying this open reading frame might have simply altered the expression level of the nearby erf gene (Semerjian et al., 1989). Nonetheless, P22 phage containing either an in-frame deletion or an amber mutation in the arf gene showed a 4- to 5-fold decrease in the rate of recombination in P22 phage crosses, a defect that could be complemented by plasmids expressing the arf gene in trans (Poteete et al., 1991). Given the highly acidic nature of the Arf protein, it was suggested that it might play a role as a DNA mimic, perhaps helping displace Erf from ssDNA during the annealing reaction. In this role, the Arf protein might not be specific for the phage recombination system, as its overexpression from a plasmid also stimulated a small (2.5-fold) effect in λ red gam phage recombination promoted by the host recombination system (Poteete et al., 1991).

The P22 recombination system could promote recombination of its chromosome by any of the mechanisms discussed below for λ Red. This view is supported by the fact that the two recombination systems can substitute for one another (Poteete and Fenton, 1984). However, P22 has the added feature that recombination is essential for its life cycle, as circularization of its chromosome following infection is a prerequisite for growth (presumably via an ssDNA annealing mechanism between it terminally repetitious ends). It is interesting to speculate that this feature of the P22 life cycle might have dictated the greater complexity of its recombination system relative to phage λ (four P22 genes relative to two λ genes), providing a level of regulation that ensures recombination will take place consistently soon after infection. A summary of the recombination functions for phages lambda, rac and P22 is listed in Table I.

Table I. Summary of phage recombination functions

Roles of VirD2 Relaxase in T-DNA Processing and Transfer

It is now widely accepted that DNA-processing reactions associated with T-DNA transfer are equivalent to those mediating bacterial conjugation. In the generalized reaction, a set of proteins termed as the DNA transfer and replication (Dtr) proteins assemble at an origin-of-transfer (oriT) sequence to generate a nucleoprotein complex termed as the relaxosome. One component of the relaxosome, the relaxase, cleaves and remains covalently associated with the 5′ end of the DNA strand destined for transfer (T-strand). The T-strand is unwound from its template by a strand displacement reaction, generating the translocation-competent relaxase-T-strand substrate. In A. tumefaciens, the VirD2 relaxase generates nicks at oriT-like sequences located in the T-DNA border repeats. VirD2 remains covalently bound to the 5′ phosphoryl end of the nicked T-DNA via conserved tyrosine residue Tyr-29. Purified VirD2 catalyzes cleavage of oligonucleotides bearing a T-DNA nick site. However, an ancillary protein, VirD1, is essential for nicking in vitro when the nick site is present on a supercoiled, double-stranded plasmid.

In addition to oriT nicking, the relaxase component of the conjugative transfer intermediate is thought to participate in translocation of substrate DNA by supplying a signal motif recognizable by the transport machinery. VirD2 and other relaxases carry a motif at their extreme C termini that is devoid of secondary structure and rich in positively charged amino acids, particularly arginines. This motif is also present at the C-termini of protein substrates of the VirB/D4 T4S system and, as expected, mutations in the signal motif of one such substrate, VirF, block translocation. The charged motif likely confers recognition of the substrate by the secretion channel, as suggested by evidence that the VirD2-T-strand complex, as well as another protein substrate, VirE2, interact with the VirD4 substrate receptor (SR). Moreover, when the C-terminal fragment of VirE2 is fused to the green fluorescent protein (GFP), it mediates binding of the reporter protein to VirD4 in living cells. Early studies supplied evidence for 5′-3′ unidirectional transfer of the T-strand, which is also compatible with the notion that the relaxase serves to pilot the attached T-strand through the secretion channel.