CRISPR-Cas systems restrict horizontal gene transfer in Pseudomonas aeruginosa

CRISPR-Cas systems provide bacteria and archaea with an adaptive immune system that targets foreign DNA. However, the xenogenic nature of immunity provided by CRISPR-Cas raises the possibility that these systems may constrain horizontal gene transfer. Here we test this hypothesis in the opportunistic pathogen Pseudomonas aeruginosa, which has emerged as an important model system for understanding CRISPR-Cas function. Across the diversity of P. aeruginosa, active CRISPR-Cas systems are associated with smaller genomes and higher GC content, suggesting that CRISPR-Cas inhibits the acquisition of foreign DNA. Although phage is the major target of CRISPR-Cas spacers, more than 80% of isolates with an active CRISPR-Cas system have spacers that target integrative conjugative elements (ICE) or the conserved conjugative transfer machinery used by plasmids and ICE. Consistent with these results, genomes containing active CRISPR-Cas systems harbour a lower abundance of both prophage and ICE. Crucially, spacers in genomes with active CRISPR-Cas systems map to ICE and phage that are integrated into the chromosomes of closely related genomes lacking CRISPR-Cas immunity. We propose that CRISPR-Cas acts as an important constraint to horizontal gene transfer, and the evolutionary mechanisms that ensure its maintenance or drive its loss are key to the ability of this pathogen to adapt to new niches and stressors.Subject terms: Microbiology, Microbial genetics, Evolution     

Exploration of horizontal gene transfer between transplastomic tobacco and plant-associated bacteria

The likelihood of gene transfer from transgenic plants to bacteria is dependent on the transgene copy number and on the presence of homologous sequences for recombination. The large number of chloroplast genomes in a plant cell as well as the prokaryotic origin of the transgene may thus significantly increase the likelihood of gene transfer from transplastomic plants to bacteria. In order to assess the probability of such a transfer, bacterial isolates, screened for their ability to colonize decaying tobacco plant tissue and possessing DNA sequence similarity to the chloroplastic genes accD and rbcL flanking the transgene (aadA), were tested for their ability to take up extracellular DNA (broad host-range pBBR1MCS-3-derived plasmid, transplastomic plant DNA and PCR products containing the genes accD-aadA-rbcL) by natural or electrotransformation. The results showed that among the 16 bacterial isolates tested, six were able to accept foreign DNA and acquire the spectinomycin resistance conferred by the aadA gene on plasmid, but none of them managed to integrate transgenic DNA in their chromosome. Our results provide no indication that the theoretical gene transfer-enhancing properties of transplastomic plants cause horizontal gene transfer at rates above those found in other studies with nuclear transgenes.     

Evidence for particle-induced horizontal gene transfer and serial transduction between bacteria

Incubation of the amino acid-deficient strain Escherichia coli AB1157 with particles harvested from an oligotrophic environment revealed evidence of horizontal gene transfer (HGT) with restoration of all deficiencies in revertant cells with frequencies up to 1.94 × 10(-5). None of the markers were preferentially transferred, indicating that the DNA transfer is performed by generalized transduction. The highest gene transfer frequencies were obtained for single markers, with values up to 1.04 × 10(-2). All revertants were able to produce particles of comparable size, appearing at the beginning of the stationary phase. Examination of the revertants using electron microscopy showed bud-like structures with electron-dense bodies. The particles that display the structural features of membrane vesicles were again infectious to E. coli AB1157, producing new infectious particles able to transduce genetic information, a phenomenon termed serial transduction. Thus, the <0.2-μm particle fraction from seawater contains a particle size fraction with high potential for gene transfer. Biased sinusoidal field gel electrophoresis indicated a DNA content for the particles of 370 kbp, which was higher than that of known membrane vesicles. These findings provide evidence of a new method of HGT, in which mobilizable DNA is trafficked from donor to recipient cells via particles.     

Mechanisms of, and barriers to, horizontal gene transfer between bacteria

Bacteria evolve rapidly not only by mutation and rapid multiplication, but also by transfer of DNA, which can result in strains with beneficial mutations from more than one parent. Transformation involves the release of naked DNA followed by uptake and recombination. Homologous recombination and DNA-repair processes normally limit this to DNA from similar bacteria. However, if a gene moves onto a broad-host-range plasmid it might be able to spread without the need for recombination. There are barriers to both these processes but they reduce, rather than prevent, gene acquisition.     

Evidence for horizontal gene transfer from bacteroidetes bacteria to dinoflagellate minicircles

Dinoflagellate protists harbor a characteristic peridinin-containing plastid that evolved from a red or haptophyte alga. In contrast to typical plastids that have ～100-200 kb circular genomes, the dinoflagellate plastid genome is composed of minicircles that each encode 0-5 genes. It is commonly assumed that dinoflagellate minicircles are derived from a standard plastid genome through drastic reduction and fragmentation. However, we demonstrate that the ycf16 and ycf24 genes (encoded on the Ceratium AF490364 minicircle), as well as rpl28 and rpl33 (encoded on the Pyrocystis AF490367 minicircle), are related to sequences from Algoriphagus and/or Cytophaga bacteria belonging to the Bacteroidetes clade. Moreover, we identified a new open reading frame on the Pyrocystis minicircle encoding a SRP54 N domain, which is typical of FtsY proteins. Because neither of these minicircles share sequence similarity with any other dinoflagellate minicircles, and their genes resemble bacterial operons, we propose that these Ceratium and Pyrocystis minicircles resulted from a horizontal gene transfer (HGT) from a Bacteroidetes donor. Our findings are the first indication of HGT to dinoflagellate minicircles, highlighting yet another peculiar aspect of this plastid genome.     

Viunalikeviruses are environmentally common agents of horizontal gene transfer in pathogens and biocontrol bacteria

Bacteriophages have been used as natural biocontrol and therapeutic agents, but also as biotechnological tools for bacterial engineering. We showed recently that the transducing bacteriophage ϕMAM1 is a ViI-like phage and a member of the new genus, ‘Viunalikevirus''. Here, we show that four additional ViI-like phages and three new environmentally isolated viunalikeviruses, all infecting plant and human pathogens, are very efficient generalised transducers capable of transducing chromosomal markers at frequencies of up to 10⁻⁴ transductants per plaque-forming unit. We also demonstrate the interstrain transduction of plasmids and chromosomal markers, including genes involved in anabolism, genes for virulence and genes encoding secondary metabolites involved in biocontrol. We propose that all viunalikeviruses are likely to perform efficient horizontal gene transfer. Viunalikeviruses therefore represent useful agents for functional genomics and bacterial engineering, and for chemical and synthetic biology studies, but could be viewed as inappropriate choices for phage therapy.Combined morphological, genomic and phylogenetic analyses have recently led to the proposed creation of a new phage genus, ‘Viunalikevirus'', within the Myoviridae family (Adriaenssens et al., 2012a). The first member of this proposed genus, Salmonella phage ViI, was isolated in the 1930s (Craigie and Yen, 1938) and multiple viunalikeviruses have been sequenced and characterised since 2010 (Pickard et al., 2010; Anany et al., 2011; Hooton et al., 2011; Kutter et al., 2011; Matilla and Salmond, 2012; Park et al., 2012; Adriaenssens et al., 2012a, 2012b; Hsu et al., 2013; Luna et al., 2013; Shahrbabak et al., 2013). Viunalikeviruses are characterised as virulent (lytic) phages showing similar genome size, extensive DNA homology, strong gene synteny and a complex adsorption apparatus, which uses tail spike proteins as host-recognition determinants (Adriaenssens et al., 2012a).We recently isolated the ViI-like phage, ϕMAM1, that infects several environmental and clinical isolates belonging to Serratia and Kluyvera genera (Matilla and Salmond, 2012). During the characterisation of ϕMAM1, we showed that it mediates highly efficient generalised transduction (Matilla and Salmond, submitted for publication). These observations were consistent with a previous report, that the Salmonella phage ViI was also capable of transduction (Cerquetti and Hooke, 1993) and we have confirmed that phage ViI can transduce chromosomal markers and plasmids at frequencies of up to 4.6 × 10⁻⁵ transductants per plaque-forming unit (p.f.u.; Figure 1a; Supplementary Table 1).Open in a separate windowFigure 1Transduction capabilities of viunalikeviruses. (a) Transduction frequencies of LIMEstone1, LIMEstone2, ViI and CBA120 phages. The graph also shows transduction efficiencies of LIMEstone phages within and between Dickeya solani strains. Transduction efficiency was defined as the number of transductants obtained per p.f.u. In all cases, error bars represent the standard deviations (n=3). (b) Skimmed milk agar plates showing protease production in the wild-type (wt) Dickeya solani strains MK10, MK16 and IPO 2222. LIMEstone1- (LS1) and LIMEstone2- (LS2) mediated transduction of the spp::Km marker from the protease negative mutant strain MK10P1 to the wild-type strains MK10, MK16 and IPO 2222 result in a protease-negative phenotype. (c–e) LIMEstone-mediated transduction of the oocN::Km marker from the oocydin A-negative mutant strain MK10oocN to the wild-type strains MK10 (c), MK16 (d) and IPO 2222 (e) results in an oocydin A-negative phenotype and, consequently, in the generation of strains defective in their antimicrobial activity against the plant pathogenic oomycete, Pythium ultimum. The anti-oomycete assays were performed as described previously (Matilla et al., 2012).Most generalised transducers utilise a headful packing strategy where phage terminases recognise specific sequences (pac sites) in the DNA and perform cycles of packing that result in mature phage particles (Fineran et al., 2009a). Indeed, phage terminases with reduced specificity for pac sequences may lead to the evolution of efficient transducing phages (Schmeiger, 1972). Based on the high similarity between the terminases of ϕMAM1, ViI and those of other previously sequenced viunalikeviruses, we hypothesised that all of these ViI-like phages should be capable of transduction in their respective bacterial hosts. To test this hypothesis, we investigated three additional viunalikeviruses, Escherichia coli phage CBA120 (Kutter et al., 2011), and Dickeya phages LIMEstone1 and LIMEstone2 (Adriaenssens et al., 2012b). All the bacteriophages, bacterial strains, plasmids and primers used in this study are listed in the Supplementary Tables 2 and 3. Experimental procedures are presented as Supplementary Material.The LIMEstone phages specifically infect some strains of the emerging plant pathogen, Dickeya solani (Adriaenssens et al., 2012b), and here we showed that they also infect the recently sequenced D. solani strains MK10, MK16 and IPO 2222. As predicted, we confirmed that the LIMEstone phages effected efficient transduction of various auxotrophic markers between Dickeya solani strains (Figure 1a; Supplementary Table 4). To our knowledge, only one Dickeya transducing phage, ϕEC2, has been isolated previously (Resibois et al., 1984). Additional mutant strains were constructed and the generalised nature of the transduction was confirmed by transfer of multiple chromosomal markers, including mutations in the gene cluster encoding biosynthesis of the anti-oomycete haterumalide, oocydin A (Matilla et al., 2012) and in the locus for synthesis and secretion of protease virulence factors. Transduction frequency was higher at an multiplicity of infection (m.o.i.) of 0.1 and 0.01 with efficiencies of up to 10⁻⁴ transductants per p.f.u. (Figure 1a; Supplementary Tables 4 and 5).We also demonstrated transduction of a kanamycin resistance-marked plasmid pECA1039-Km3 between strains MK10, MK16 and IPO 2222 at frequencies of up to 8.6 × 10⁻⁵ (Supplementary Table 4). Plasmid pECA1039 (originally isolated from the phytopathogen, Pectobacterium atrosepticum) encodes a bifunctional type III Toxin-Antitoxin (TA) system, ToxIN, with abortive infection capacity. Although ToxIN aborts infection of various enterobacteria by diverse phages (Fineran et al., 2009b) it did not protect against infection by the tested viunalikeviruses, ϕMAM1, ViI, CBA120, LIMEstone1 or LIMEstone2 (not shown). Furthermore, another type III TA system, TenpIN, from the insect pathogen, Photorhabdus luminescens (Blower et al., 2012), failed to protect against any of the five ViI-like phages (not shown).In addition, we also tested the transduction capacity of the E. coli phage, CBA120, and confirmed transduction of plasmid-borne antibiotic resistances at a frequency of up to 10⁻⁴ transductants per p.f.u. (Figure 1a; Supplementary Table 6).We decided to test our hypothesis that the viunalikeviruses may all be generalised transducers by first isolating new viunalikeviruses from the environment. From treated sewage effluent, we isolated three new bacteriophages infecting Dickeya solani, ϕXF1, ϕXF3 and ϕXF4, as defined initially by their very characteristic ViI-like morphology in electron microscopy (Figures 2a–c). As predicted, all of these new phages were able to transduce chromosomal markers and plasmids at frequencies of up to 3 × 10⁻⁶ transductants per p.f.u. (Figure 2e; Supplementary Table 7). Sequencing of structural and non-structural protein-encoding genes of ϕXF1, ϕXF3 and ϕXF4 showed high nucleotide homology (between 80% and 100%) with the corresponding orthologs in LIMEstone1 (Supplementary Figure 1), indicating that these virgin environmental isolates also clade within the Viunalikevirus genus.Open in a separate windowFigure 2Environmental isolation and characterisation of new viunalikeviruses with generalised transduction functionality. Transmission electron micrographs of phages ϕXF1 (a), ϕXF3 (b), ϕXF4 (c) and ϕXF28 (d) are shown. As an internal control, ϕXF28 was an example of a new lytic phage isolated from the same environment but showing no transduction capabilities. Bars, 50 nm. (e) Transduction frequencies of the new viunalikeviruses ϕXF1, ϕXF3 and ϕXF4. Transduction experiments were performed using 10⁹ cells with ϕXF1, ϕXF3, ϕXF4 at an m.o.i. of 0.01. Transduction efficiency was defined as the number of transductants obtained per p.f.u. Error bars represent the standard deviations (n=3).Although we did not have access to other ViI-like Salmonella phages SFP10 (Park et al., 2012), ϕSH19 (Hooton et al., 2011) and Marshall (Luna et al., 2013), Escherichia phage PhaxI (Shahrbabak et al., 2013), Shigella phage ϕSboM-AG3 (Anany et al., 2011) and Klebsiella phage 0507-KN2-1 (Hsu et al., 2013), our results allow us to predict that all of these phages will mediate generalised transduction. Importantly, these phages would be expected to contribute to the horizontal gene transfer of virulence factors and antimicrobial-resistance determinants in diverse environments.Viunalikeviruses do not seem to be limited to the enterobacteria as bacteriophages showing ViI-like morphology have been isolated in Acinetobacter (Ackermann et al., 1994), Bordetella (Adriaenssens et al., 2012b) and Sinorhizobium (Werquin et al., 1988). Furthermore, another ViI-like morphotype phage (ϕM12 of Sinorhizobium meliloti) has also been shown to be an efficient transducer (Finan et al., 1984). Taken together, these results suggest that, even in the absence of strongly predictive comparative genomic detail, a characteristically discrete ViI-like morphology in electron microscopy may be sufficient to identify new phages as strong candidates for possession of generalised transduction capacity.The emergence and dissemination of antibiotic-resistant pathogens coupled with low discovery rates for new antimicrobials, plus increasing legal constraints on the use of chemical pesticides, have (re)focussed attention on the potential use of bacteriophages for ‘natural biocontrol'' of human, animal and plant pathogens. Several viunalikeviruses have been proposed as candidate therapeutic agents for the control of bacterial infections (Anany et al., 2011; Hooton et al., 2011; Park et al., 2012; Hsu et al., 2013; Shahrbabak et al., 2013) and the LIMEstone phages have been used in successful field trials for biocontrol of D. solani infections (Adriaenssens et al., 2012b). However, their efficient transduction capacities could provide a route for dissemination of virulence factors, such as proteases (Marits et al., 1999). In fact, we have demonstrated the interstrain transduction of plasmids and oocydin A, auxotrophy and protease markers between three different D. solani strains, at high frequencies (Figures 1 and ​and2;2; Supplementary Tables 4 and 7). Also, the irregular distribution of the oocydin A gene cluster within the Dickeya genus and the fact that its genomic context varies between strains raises the possibility of phage-mediated horizontal gene transfer between bacterial strains. These results emphasize strongly that when considering the genomics of phages for ‘phage therapy'' the absence of genes readily defined as playing roles in lysogeny or bacterial virulence may be insufficient to inspire confidence that use of a particular therapeutic phage presents no risk–particularly among the high efficiency-transducing viunalikeviruses.     

Role of horizontal gene transfer by bacteriophages in the origin of pathogenic bacteria

The review considers the involvement of bacteriophages in transferring genes, which determine bacterial pathogenicity, and the increasing role of comparative genomics and genetics of bacteria and bacteriophages in detecting new cases of horizontal gene transfer. Examples of phage participation in this process proved to a different extent are described. Emphasis is placed on the original work carried out in Russia and focused on bacteriophages (temperate transposable phages and giant virulent phi KZ-like phages) of conditional pathogen Pseudomonas aeruginosa. Consideration is given to the possible lines of further research of the role of bacteriophages in the infection process and, in particular, the role of virulent phages, whose products are similar to those of pathogenic bacteria, in modification of clinical signs of infectious diseases and in evolution. An attempt is made to predict the possible direction of pathogen evolution associated with development of new treatment strategies and generation of new specific niches.     

Comparative analysis of magnetosome gene clusters in magnetotactic bacteria provides further evidence for horizontal gene transfer

The organization of magnetosome genes was analysed in all available complete or partial genomic sequences of magnetotactic bacteria (MTB), including the magnetosome island (MAI) of the magnetotactic marine vibrio strain MV‐1 determined in this study. The MAI was found to differ in gene content and organization between Magnetospirillum species and strains MV‐1 or MC‐1. Although a similar organization of magnetosome genes was found in all MTB, distinct variations in gene order and sequence similarity were uncovered that may account for the observed diversity of biomineralization, cell biology and magnetotaxis found in various MTB. While several magnetosome genes were present in all MTB, others were confined to Magnetospirillum species, indicating that the minimal set of genes required for magnetosome biomineralization might be smaller than previously suggested. A number of novel candidate genes were implicated in magnetosome formation by gene cluster comparison. Based on phylogenetic and compositional evidence we present a model for the evolution of magnetotaxis within the Alphaproteobacteria, which suggests the independent horizontal transfer of magnetosome genes from an unknown ancestor of magnetospirilla into strains MC‐1 and MV‐1.     

Monitoring and modeling horizontal gene transfer

Monitoring efforts have failed to identify horizontal gene transfer (HGT) events occurring from transgenic plants into bacterial communities in soil or intestinal environments. The lack of such observations is frequently cited in biosafety literature and by regulatory risk assessment. Our analysis of the sensitivity of current monitoring efforts shows that studies to date have examined potential HGT events occurring in less than 2 g of sample material, when combined. Moreover, a population genetic model predicts that rare bacterial transformants acquiring transgenes require years of growth to out-compete wild-type bacteria. Time of sampling is there-fore crucial to the useful implementation of monitoring. A population genetic approach is advocated for elucidating the necessary sample sizes and times of sampling for monitoring HGT into large bacterial populations. Major changes in current monitoring approaches are needed, including explicit consideration of the population size of exposed bacteria, the bacterial generation time, the strength of selection acting on the transgene-carrying bacteria, and the sample size necessary to verify or falsify the HGT hypotheses tested.     

Eukaryotic signalling domain homologues in archaea and bacteria. Ancient ancestry and horizontal gene transfer.

Phyletic distributions of eukaryotic signalling domains were studied using recently developed sensitive methods for protein sequence analysis, with an emphasis on the detection and accurate enumeration of homologues in bacteria and archaea. A major difference was found between the distributions of enzyme families that are typically found in all three divisions of cellular life and non-enzymatic domain families that are usually eukaryote-specific. Previously undetected bacterial homologues were identified for# plant pathogenesis-related proteins, Pad1, von Willebrand factor type A, src homology 3 and YWTD repeat-containing domains. Comparisons of the domain distributions in eukaryotes and prokaryotes enabled distinctions to be made between the domains originating prior to the last common ancestor of all known life forms and those apparently originating as consequences of horizontal gene transfer events. A number of transfers of signalling domains from eukaryotes to bacteria were confidently identified, in contrast to only a single case of apparent transfer from eukaryotes to archaea.     

Organization of butyrate synthetic genes in human colonic bacteria: phylogenetic conservation and horizontal gene transfer

Butyrate producers constitute an important bacterial group in the human large intestine. Butyryl-CoA is formed from two molecules of acetyl-CoA in a process resembling beta-oxidation in reverse. Three different arrangements of the six genes coding for this pathway have been found in low mol% G+C-content gram-positive human colonic bacteria using DNA sequencing and degenerate PCR. Gene arrangements were strongly conserved within phylogenetic groups defined by 16S rRNA gene sequence relationships. In the case of one of the genes, encoding beta-hydroxybutyryl-CoA dehydrogenase, however, sequence relationships were strongly suggestive of horizontal gene transfer between lineages. The newly identified gene for butyryl-CoA CoA-transferase, which performs the final step in butyrate formation in most known human colonic bacteria, was not closely linked to these central pathway genes.     

Type IV secretion systems: tools of bacterial horizontal gene transfer and virulence

Type IV secretion systems (T4SSs) are multisubunit cell-envelope-spanning structures, ancestrally related to bacterial conjugation machines, which transfer proteins and nucleoprotein complexes across membranes. T4SSs mediate horizontal gene transfer, thus contributing to genome plasticity and the evolution of pathogens through dissemination of antibiotic resistance and virulence genes. Moreover, T4SSs are also used for the delivery of bacterial effector proteins across the bacterial membrane and the plasmatic membrane of eukaryotic host cell, thus contributing directly to pathogenicity. T4SSs are usually encoded by multiple genes organized into a single functional unit. Based on a number of features, the organization of genetic determinants, shared homologies and evolutionary relationships, T4SSs have been divided into several groups. Type F and P (type IVA) T4SSs resembling the archetypal VirB/VirD4 system of Agrobacterium tumefaciens are considered to be the paradigm of type IV secretion, while type I (type IVB) T4SSs are found in intracellular bacterial pathogens, Legionella pneumophila and Coxiella burnetii. Several novel T4SSs have been identified recently and their functions await investigation. The most recently described GI type T4SSs play a key role in the horizontal transfer of a wide variety of genomic islands derived from a broad spectrum of bacterial strains.     

Characterization of mimivirus DNA topoisomerase IB suggests horizontal gene transfer between eukaryal viruses and bacteria

Mimivirus, a parasite of Acanthamoeba polyphaga, is the largest DNA virus known; it encodes dozens of proteins with imputed functions in nucleic acid transactions. Here we produced, purified, and characterized mimivirus DNA topoisomerase IB (TopIB), which we find to be a structural and functional homolog of poxvirus TopIB and the poxvirus-like topoisomerases discovered recently in bacteria. Arginine, histidine, and tyrosine side chains responsible for TopIB transesterification are conserved and essential in mimivirus TopIB. Moreover, mimivirus TopIB is capable of incising duplex DNA at the 5'-CCCTT cleavage site recognized by all poxvirus topoisomerases. Based on the available data, mimivirus TopIB appears functionally more akin to poxvirus TopIB than bacterial TopIB, despite its greater primary structure similarity to the bacterial TopIB group. We speculate that the ancestral bacterial/viral TopIB was disseminated by horizontal gene transfer within amoebae, which are permissive hosts for either intracellular growth or persistence of many present-day bacterial species that have a type IB topoisomerase.     

Detecting highways of horizontal gene transfer

In a horizontal gene transfer (HGT) event, a gene is transferred between two species that do not have an ancestor-descendant relationship. Typically, no more than a few genes are horizontally transferred between any two species. However, several studies identified pairs of species between which many different genes were horizontally transferred. Such a pair is said to be linked by a highway of gene sharing. We present a method for inferring such highways. Our method is based on the fact that the evolutionary histories of horizontally transferred genes disagree with the corresponding species phylogeny. Specifically, given a set of gene trees and a trusted rooted species tree, each gene tree is first decomposed into its constituent quartet trees and the quartets that are inconsistent with the species tree are identified. Our method finds a pair of species such that a highway between them explains the largest (normalized) fraction of inconsistent quartets. For a problem on n species and m input quartet trees, we give an efficient O(m + n(2))-time algorithm for detecting highways, which is optimal with respect to the quartets input size. An application of our method to a dataset of 1128 genes from 11 cyanobacterial species, as well as to simulated datasets, illustrates the efficacy of our method.