Genomic islands tools for

A number of GEIs are capable of integration into the chromosome of the host, excision, and transfer to a new host by transformation, conjugation or transduction. GEIs play a crucial role in the evolution of a broad spectrum of bacteria as they are involved in the dissemination of variable genes, including antibiotic resistance and virulence genes leading to generation of hospital 'superbugs', as well as catabolic genes leading to formation of new metabolic pathways.

Depending on the composition of gene modules, the same type of GEIs can promote survival of pathogenic as well as environmental bacteria.

PAIs, for example, can cause major changes in the bacterial phenotype. Thus, they are the most studied GIs Hacker and Carniel, The ability of bacteria to transmit pathogenicity factors and antibiotic resistance factors is one of the most widely studied topics associated with GIs.

The high prevalence of antibiotic resistance is an important problem facing the health care system, as it jeopardizes the success of treating infectious diseases. Changes in bacterial populations, which have increased their resistance level to various antibiotics within a few decades, show that bacteria adapt and evolve rapidly. GIs are associated with an increased distribution of virulence and antibiotic resistance factors, indicating their importance in the evolution of bacterial genomes Juhas et al.

The large number of sequenced genomes and analyses of genetic sequences have revealed that GIs are mosaics of genes formed by HGT. Several methods for GI prediction and genomic data analysis have been developed. The main methods used by prediction tools are separated into two groups: comparative genomic analysis, whose objective is to identify variable regions in relatively close organisms multiple genomes , and analysis of sequence composition in the organism single genome Lu and Leong, b.

Although numerous prediction tools are available, the accuracy of the results is insufficient. The use of only one method may not give satisfactory results; the combination of various techniques may be a better strategy for bridging the gaps in genomic island prediction Lu and Leong, b.

Recently, Bertelli et al. Analysis of the methods applied in each tool provided a broad view of the applicability of each software, revealing which predictors are better for the data set.

Based on the results, Soares et al. Our objective was to qualitatively and quantitatively evaluate these prediction tools against manually curated GIs. We used a set of diverse organisms and known islands curated in vitro to evaluate the prediction methods, island behavior in different organisms, and processes of adaptation and genomic evolution. We chose the following predictors that use sequence composition: Alien Hunter Vernikos and Parkhill, , Predict Bias Pundhir et al.

GIPSy performs analysis using the methods incorporated in the Artemis genome visualization tool Rutherford et al. Predict Bias uses GenBank files to identify tRNAs, transposases, and integrases and determines the relationship between island function and pathogenicity.

The various methods and integrated tools used by the chosen predictors to identify the main characteristics of the GIs provides a broad view of the results for analyzing and comparing the same dataset to determine which tools give the best results. Table 1 describes the chosen tools and their main characteristics. A complete description of the prediction tools is shown in Supplementary Table 1.

To evaluate these characteristics, we searched the literature for in vitro curated GIs that are already well-defined. The predictors were excluded based on 1 low performance according to a previous study Bertelli et al. The tools evaluated and excluded are shown in Supplementary Table 2. The organisms chosen as the test set for this study were selected from those used in previous studies describing the tools; all chosen organism-genomes had been tested by at least two other tools.

We tested only full genomes because not all organisms have additional information available, such as plasmids and viruses. We selected bacteria from different families to ensure diversity in our analysis. Of these bacteria, three were gram-positive and five were gram-negative Table 2.

Using GIs previously analyzed and used as in vitro as reference data enabled us to evaluate the sensitivity and accuracy of the tools. The authors Lloyd et al. Additionally, the biological composition of the GIs described and identified in vitro was consistent with several analytical features present in the chosen predictors.

Several GIs of the gold standard have well-defined tRNA and PAI functions, enabling comparative analysis of the predictor results with curated databases for these specific characteristics. IslandViewer4 implemented Islander in its last update, but only the pre-computed results are available for consultation users uploading their own genome cannot receive Islander results. Because of the lack of data in the reference banks when assimilated with the data from the literature, we manually curated all predictors results, identifying each biological product found by the tools and relating it to their functions and characteristics.

Table 3 shows the data from the in vitro curated islands of E. Table 3. The gold standard GIs are represented by the first locus tag of the region and last locus tag of the region in the genome see Table 3. We performed locus tag conversion to compare the results because the GIs predicted by the tools were identified by the initial and final position of the candidate GIs in the genome.

Because the tools used different methods, the positions of the predicted islands may not be exact compared to the positions of the standard GIs, both for the beginning and end of the island. Curation was performed manually using the results from each tool.

From the results of candidate islands predicted by the different tools, we generated. GBK files for each island from Artemis software. Unique GIs were identified by only one predictor.

The flowchart in Figure 2 shows the steps used for dataset construction. The Venn diagrams were examined in detail using the web tool InteractiVenn Heberle et al. The results of the Intersection Plot Graph and Venn diagram are the data sets of common and unique GIs by organism and the total data sets compared to all predictors against each other to identify similarity hits between them.

Table 4 shows the main features of each tool and Table 5 shows the complementary information. Developed by researchers at the Sanger Institute in the UK.

The predictions can be optimized using two-state Hidden Markov Models HMM to identify the entry point in the atypical and non-atypical regions of the genome Vernikos and Parkhill, When the identification of these regions occurs, IVOM score is obtained, which is equivalent to how much this portion of the genome differs from the rest.

Longer sequences have higher scores and more accurate predictions, whereas smaller sequences with few information have a lower score and a questionable result Che et al. Threshold is also established with a score, resulting from the comparison with the average of the total genome related to its similarity. Genes or genomic regions with a score below or above the threshold are possibly atypical, subsequent genes or even these atypical regions are linked to obtain candidates GIs Lu and Leong, b.

Alien Hunter is able to make predictions without requiring a pre-existing annotation. Therefore, it can be used in newly sequenced genomes Che et al. It is able to identify GIs in both bacterial and archaea genomes.

It is based on analyzes of sequence composition, tRNA genes and highly expressed genes, intergenic distance, information on phages, and mobile genes integrase and transposases , as well as the implementation of the Interpolated Variable Order Motif IVOM methodology that the Alien tool Hunter uses it to perform analyzes Che et al.

In order to predict the GIs, a decision tree based prediction method with a training set was also developed. The attributes of the highly expressed genes and the intergenic distances were not explored in other tools Che et al.

After improving, GIPSy is able to identify other candidate regions, as well as classify them according to the genes present in the GIs in relation to their biological functions Mls, Rls, Sls. To perform the analyzes a reference genome is required.

It is also a database of GIs containing bacterial and archaea organisms. The interactive genome graph is provide in the web page, which gives the user a broad view of all predicted GIs with their products and features; indicates the genes related to virulence factors, pathogenicity, and antibiotic resistance. This tool does not allow the user to choose the reference genome for the IslandPick method before making the prediction.

Only after receiving the results can the user choose another related genome for comparison Bertelli et al. This predictor identifies genomic and pathogenic islands in prokaryotic organisms from the evaluation of sequence composition, presence of insertion elements and genes related to virulence factors. It is a non-supervised and algorithm-dependent annotation tool for automated targeting.

This approach combines homogeneity of sequences within each island and heterogeneity of sequence compositions. Figure 3 shows the performance of the evaluated predictors with respect to the processing time. Figure 3. Organism with the larger genome, Pseudomonas aeruginosa PAO1 6,, base pairs was compared to that with a smaller genome, Streptococcus pyogenes M1 GAS 1,, base pairs.

The processing time of Predict Bias was not influenced by the genome size. Because this tool uses a set of databases, we hypothesized that some annotations had been preprocessed.

The processing time of Zisland Explorer was slightly influenced by the size of the analyzed genomes. We did not analyze unmarked genomes.

Therefore, we cannot infer an estimated time for these type of predictions. However, this software uses two genomes for analysis study and reference , and thus its runtime may vary. The broadband does not appear to directly influence the time required for IslandViewer4 to perform the analyses.

However, this information is not included in the published articles or on the tool page. The tool uses several processes in its analyses. The time may be influenced by the number of queries being processed at a specific time in relation to queries from other organisms previously sent by other users.

In conclusion, all tools showed a relatively fast runtime, and none presented errors during execution. We evaluated which tool most closely predicted the 16 GIs curated in vitro Lloyd et al. Figure 4 shows the positions of the 16 in vitro curated GIs on the genome plotted by Artemis and the predictors used for identification.

Figure 4. No predictors matched the 16 GIs previously reported for the gold standard, but each island was predicted by one or more tools. Only one island was identified by all tools GI Table 6 shows a summary of the GI 16 content. It is not associated with tRNA genes, and has one integrase and one transposase. Alien Hunter and IslandViewer4 showed the best predictions. Both tools identified the mobility genes present on the island. IslandViewer4 failed to identify tRNAs or integrase present in the island.

GIPSy identified all mobility genes, and IslandViewer4 was associated one integrase and two transposases. GIPSy showed the best results. According to in vitro curation, this island lacks a tRNA Lloyd et al. Both tools identified the three tRNA genes together with the transposase but failed to identify the integrase.

Both tools identified the integrase gene present on the island. PAI 9 contains the fyuA gene encoding a yersiniabactin receptor, a siderophore found in pathogenic bacteria. FyuA is important for biofilm formation in disadvantageous environments with high contents of iron, such as in human urine Hancock et al. A transposase lies in the middle of the island and fyuA is at the end.

A threshold was used for identifying atypical regions in the genome; for this prediction, the threshold was PAI 10 contains the tcpC gene, which is responsible for interfering with the innate immune response of the host Erjavec et al.

The tcpC gene is found in the middle of the island. However, in the GBK annotation, this gene was marked as a hypothetical protein. GI 12 predominantly contains bacteriophage DNA. This island has no tRNA genes and only one integrase. IslandViewer4 identified the entire region and its CDS. Table 7 shows the relevant products of the 16 GIs of the gold standard according to Lloyd et al.

Table 7. Products of the 16 gold standard gis of Escherichia coli cft based on reference articles. To compare the total results of each predictor, a survey of the 16 GIs in the gold standard was performed considering the main products such as tRNAs, integrases, transposases, hypothetical, and uncharacterized proteins, and the number of CDS in the region. We included all protein products in the CDS count. No tool presented exact predictions of the initial and final GI positions compared to the gold standard.

Some predictions lost CDS, while others included other genetic components. To guarantee that the sum result did not affect the total gene count, any island identified by the predictors containing additional CDS or any evaluated product compared to the gold standard was excluded from the final count.

Table 8 shows the total number of relevant CDS present in the 16 GIs of the gold standard compared to the total results of the predictors. Table 8. Their importance is overtly manifested by the rapid dissemination of antibiotic resistance worldwide among H. The recent acquisition of antibiotic resistance genes and their rapid global spread over the past 30—40 years is an example of how GEIs contribute to bacterial diversification and adaptation. Staphylococcus aureus is a potentially pathogenic bacterium that plays a role in a broad spectrum of diseases and is a major cause of hospital-acquired infections worldwide.

Methicillin-resistant S. This antibiotic resistance is facilitated mostly by genes located on a GEI-designated staphylococcal cassette chromosome mec SCC mec Katayama et al. The five SCC mec subtypes identified so far range in size from 20 to 70 kb and confer resistance to methicillin, kanamycin, tobramycin, bleomycin, penicillins, heavy metals, tetracycline, macrolide, lincosamide and streptogramin Ito et al.

GEI SCC mec is interesting in a sense that it does not contain phage-related and tra genes or transposases and is transferred between bacteria with the help of two site-specific recombinases that catalyze its chromosomal excision and reintegration Ito et al. The origin of GEI SCC mec remains to be elucidated; however, it is hypothesized that it could originate from other staphylococci, namely Staphylococcus sciuri or Staphylococcus epidermidis.

This is suggested by the high-amino acid sequence similarities of the methicillin resistance gene mecA products of S. Furthermore, S. Enterococcus faecalis is also one of the leading agents of nosocomial infections of surgical sites, the urinary tract and bloodstream Richards et al.

Most of the virulent strains of E. Recently, the horizontal transfer of the E. This has many implications for the evolution and diversity of E.

Results from recent studies have shown that GEIs also played a key role in the evolution of pathogenic human and mammalian Mycobacterium spp. Gutierrez et al.

Speciation events in ancestral Mycobacterium spp. Several Mycobacterium tuberculosis GEIs harbour genes that have been identified previously as virulence genes in other bacteria Pethe et al. A good example illustrating the role played by the horizontally acquired GEIs in the evolution of Mycobacterium spp. Siderophore-mediated iron uptake is important for pathogenic as well as environmental bacteria.

This GEI was first found in Yersinia spp. Many bacterial species exploit specialized secretion systems to transfer macromolecules across bacterial membranes. GEIs of a wide variety of bacterial pathogens encode type III secretion systems T3SS and T4SS, which by transfer of proteins or nucleoprotein complexes directly mediate pathogenicity and horizontal gene transfer.

Salmonella enterica serovar Typhi is a gram-negative facultative intracellular pathogen that is a causative agent of gastroenteritis and typhoid fever Shea et al. The divergence of Salmonella and E. The first stage of Salmonella infection, characterized by the colonization and invasion of intestinal epithelial cells, is usually followed by the replication within host's macrophages. Recent findings with a mouse model of typhoid indicate that this process is even more complicated, as SPI-2 was shown to be expressed also during early stages of pathogenesis before penetrating the intestine.

This suggests that in addition to other functions, SPI-2 may be also involved in preparing Salmonella to successfully resist the harsh antimicrobial environment within macrophages Brown et al.

The T4SSs are unique among other bacterial secretion systems due to their ability to transfer both proteins and nucleoprotein complexes. They can deliver bacterial effector proteins to host cells, thus contributing directly to pathogenicity.

Furthermore, they can mediate horizontal gene transfer, thus facilitating the evolution of pathogens through dissemination of virulence genes Juhas et al. Adhesion, either interbacterial or to specific receptors of host cells, represents another important pathogenicity trait that is often associated with GEIs.

TCP is a type IV pilus that mediates interbacterial adherence, resulting in formation of microcolonies, as well as secretion of a soluble colonization factor, TcpF, crucial for successful colonization of host Kirn et al. This study suggests that independent infections of similar but distinct bacteriophages carrying virulence determinants, including LEE GEI, are deeply involved in the evolution of EHEC strains belonging to different E. Horizontal transfer of GEIs was also shown to play a major role in the evolution of a number of other bacterial pathogens, including LIPI listeria pathogenicity island of Listeria monocytogenes Vazquez-Boland et al.

Even dehalogenases in Dehalococcoides ethenogenes have been associated with integrated DNA elements, but whether these are mobile or not remains to be determined Seshadri et al. Various classical experiments have used Pseudomonas sp.

First of all, ICE clc carries a 9. Finally, in the groundwater isolate Ralstonia sp. This region was therefore suspected to contain the information for self-transfer, which was more recently confirmed by the analysis of the T4SS genes of the ICE Hin from H.

The core region of this group of elements was found in a variety of other bacteria Fig. More recently, GEI fragments similar to ICE clc were also detected in Xylella fastidiosa, Xanthomonas campestris, Rubrivivax gelatinosus, Azoarcus, Cupriviadus and the arsenic-oxidizing bacterium Herminiimonas arsenicoxydans Gaillard et al.

Various functions encoded by GEIs of the same family. Gene modules homologous in all shown GEIs are represented by white boxes, and modules specific for individual GEIs by black boxes. B13 bacteria. ATB resist. The fact that this genetic load is found mostly between the integrase gene near one end of the element and the core region suggests that insertions here are selectively neutral and do not compromise the functionality of the GEI.

A majority of accessory genes were also found to be located in this position in the closely related Haemophilus spp. Various other recent examples of whole genome comparisons show that GEIs have been implicated in adaptation to a wide range of different life-styles and carry a large variety of functions. For example, the Burkholderia cenocepacia isolates often harbour an unstable Recent comparison of several P. The dit island encodes a full metabolic pathway for abietane diterpenoids Mathee et al.

In addition, the RGP5 element of P. Magnetospirillum gryphiswaldense contains a kb unstable region of probable ancient GEI origin, but now filled with insertion elements 42 copies that cause frequent recombinations and deletions of parts of this genomic region. Interestingly, this region encodes many genes involved in magnetosome biomineralization Ullrich et al.

Suspected ancient GEIs with deteriorated integrases for which no transfer or excision could be demonstrated were detected in the anaerobic bacterium Geobacter sulfurreducens Butler et al. One such region has a size of kb and contained many genes implicated in anaerobic metabolism of benzoate, phenol, p -cresol and 4-hydroxybenzoate. Free-living water and soil-borne bacteria with no specific pathogenicity characteristics, but which are thought to provide reservoirs for more pathogenic bacteria, have now also been found to harbour GEIs.

One of those is Arcobacter butzleri , an Epsilonproteobacterium related to H. The island encodes 29 genes, none of which, however, have known orthologues in other bacterial genomes. More recent genome comparisons of multiple X. For example, Xylella strains from citrus fruits carried one or more copies of the GI1 island, which encodes fimbrillin synthesis, haemolysin production and lipopolysaccharide synthesis da Silva et al.

Various rearrangements were observed in the largest kb island of X. Work over a number of years has shown the importance that GEIs have and have had in promoting horizontal gene transfer and distributing a wide range of adaptive functions for the host bacterium. GEIs seem to do the same job as many self-transferable plasmids. Conceptually, GEIs may have a number of advantages over a plasmid, one of the most notable being that GEIs are integrated in the host's chromosome.

However, under certain conditions, those GEIs that are functionally mobile can undergo excision, self-transfer and reintegration into a new host. More interestingly, several findings, including tetracycline-induced transfer of the conjugative transposon CTn Dot of Bacteroides Cheng et al. Thus, the question arises whether unwittingly, by changing the conditions for bacteria in hospitals antibiotic stress and environment pollution , we are generating selective conditions which promote the success of self-transferable and responsive GEIs.

As shown on multiple examples in this review, GEIs play a crucial role in the evolution of a broad spectrum of pathogenic or environmental bacteria.

Furthermore, several lines of evidence suggest the existence of evolutionary ancient GEIs spread over versatile groups of otherwise unrelated bacterial species. Besides a conserved set of core genes required for maintenance, these evolutionary ancient GEIs also harbour a variable number of other gene modules whose composition is strongly dependent on the life-style of the particular host bacterial species, which can be either pathogenic or environmental Fig.

The important contribution of many research groups around the world, has led to GEI-facilitated horizontal gene transfer being one of the most rapidly evolving fields of microbiology research. We would like to cordially thank all scientists who contributed with their results and ideas to the rapidly advancing field of the horizontal gene transfer and apologize to those whose work was not cited due to space limitations.

Re-use of this article is permitted in accordance with the Creative Commons Deed, Attribution 2. National Center for Biotechnology Information , U. Fems Microbiology Reviews. Published online Oct Author information Article notes Copyright and License information Disclaimer. Published by Blackwell Publishing Ltd. This article has been cited by other articles in PMC. Abstract Bacterial genomes evolve through mutations, rearrangements or horizontal gene transfer.

Keywords: horizontal gene transfer, genomic island, evolution, pathogenicity, biodegradation. Introduction Genomes of bacterial species can evolve through a variety of processes including mutations, rearrangements or horizontal gene transfer. General features of GEIs GEIs are in essence discrete DNA segments differing between closely related bacterial strains to which usually some past or present mobility is attributed.

Open in a separate window. Evolutionary origins of GEIs Although most of the GEIs known so far fit the above-described definition, a significant number of elements lack one or more of the hallmark indications. Transfer of GEIs between bacteria As suggested above, a wide variety of GEIs are intimately connected to phages and conjugative plasmids through their evolutionary origins.

Regulation of GEIs and adaptive behaviour There is still little information available on the regulation of or environmental conditions influencing GEI transfer. Contribution of GEIs to horizontal gene transfer and bacterial evolution It is widely recognized that horizontal gene transfer facilitated by GEIs has played a crucial role in the evolution of bacterial species. Contribution of GEIs to evolution of pathogenic bacteria Antibiotic resistance represents one of the most frequent and well-studied traits associated with GEIs.

Concluding remarks Work over a number of years has shown the importance that GEIs have and have had in promoting horizontal gene transfer and distributing a wide range of adaptive functions for the host bacterium. Acknowledgments We would like to cordially thank all scientists who contributed with their results and ideas to the rapidly advancing field of the horizontal gene transfer and apologize to those whose work was not cited due to space limitations.

Statement Re-use of this article is permitted in accordance with the Creative Commons Deed, Attribution 2. Identification and characterization of the immunity repressor ImmR that controls the mobile genetic element ICEBs1 of Bacillus subtilis. Mol Microbiol. The Burkholderia cepacia epidemic strain marker is part of a novel genomic island encoding both virulence and metabolism-associated genes in Burkholderia cenocepacia.

Infect Immun. Unique features revealed by the genome sequence of Acinetobacter sp. ADP1, a versatile and naturally transformation competent bacterium. Nucleic Acids Res. Genomic and functional analyses of SXT, an integrating antibiotic resistance gene transfer element derived from Vibrio cholerae. J Bacteriol. SOS response promotes horizontal dissemination of antibiotic resistance genes. Contribution of horizontally acquired genomic islands to the evolution of the tubercle bacilli.

Mol Biol Evol. Regulatory circuits for plasmid survival. Curr Opin Microbiol. Salmonella pathogenicity island 2 is expressed prior to penetrating the intestine. PLoS Pathog. The kilobase unstable region of Yersinia pestis comprises a high-pathogenicity island linked to a pigmentation segment which undergoes internal rearrangement.

Control of SXT integration and excision. Shaping bacterial genomes with integrative and conjugative elements. Res Microbiol. Conjugative transposons: the tip of the iceberg. Genomic and microarray analysis of aromatics degradation in Geobacter metallireducens and comparison to a Geobacter isolate from a contaminated field site.

BMC Genomics. Burkholderia xenovorans LB harbors a multi-replicon, 9. The ins and outs of DNA transfer in bacteria. A selC-associated genomic island of the extraintestinal avian pathogenic Escherichia coli strain BEN is involved in carbohydrate uptake and virulence.

Biogenesis, architecture, and function of bacterial type IV secretion systems. Annu Rev Microbiol. Horizontal transfer of virulence genes encoded on the Enterococcus faecalis pathogenicity island. Genomic islands and the ecology and evolution of Prochlorococcus. Amelioration of the cost of conjugative plasmid carriage in Eschericha coli K Comparative genomic characterization of citrus-associated Xylella fastidiosa strains.

Stability of a promiscuous plasmid in different hosts: no guarantee for a long-term relationship. The molecular evolution of methicillin-resistant Staphylococcus aureus. Clin Microbiol Infect. Site-specific recombination with the chromosomal tRNA Leu gene by the large conjugative Haemophilus resistance plasmid.

Antimicrob Agents Chemother. Diversity of antibiotic resistance integrative and conjugative elements among haemophili. J Med Microbiol. Genomic islands in pathogenic and environmental microorganisms. Nat Rev Microbiol. The Salmonella genomic island 1 is an integrative mobilizable element. Evolutionary genomics of Staphylococcus aureus : insights into the origin of methicillin-resistant strains and the toxic shock syndrome epidemic. The Bacteroides fragilis pathogenicity island is contained in a putative novel conjugative transposon.

Analysis of the sequence and gene products of the transfer region of the F sex factor.