MBE Advance Access originally published online on December 27, 2007
Molecular Biology and Evolution 2008 25(3):549-558; doi:10.1093/molbev/msm282
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Articles |
A Comparative Synteny Map of Burkholderia Species Links Large-Scale Genome Rearrangements to Fine-Scale Nucleotide Variation in Prokaryotes
* Genome Institute of Singapore, Singapore
Duke-NUS Graduate Medical School, Singapore
E-mail: bourque{at}gis.a-star.edu.sg; tanbop{at}gis.a-star.edu.sg.
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Genome rearrangement events, including inversions and translocations, are frequently observed across related microbial species, but the impact of such events on functional diversity is unclear. To clarify this relationship, we compared 4 members of the Gram-negative Burkholderia family (Burkholderia pseudomallei, Burkholderia mallei, Burkholderia thailandensis, and Burkholderia cenocepacia) and identified a core set of 2,590 orthologs present in all 4 species (metagenes). The metagenes were organized into 255 synteny blocks whose relative order has been altered by a predicted minimum of 242 genome rearrangement events. Functionally, metagenes within individual synteny blocks were often related. The molecular divergence of metagenes adjacent to synteny breakpoints (boundary metagenes) was significantly greater compared with metagenes within blocks, suggesting an association between breakpoint locations and local fine-scale nucleotide alterations. This phenomenon, referred to as boundary element associated divergence, was also observed in Pseudomonas and Shigella, suggesting that this is a common phenomenon in prokaryotes. We also observed preferential localization of species-specific genes and insertion sequence element to synteny breakpoints in Burkholderia. Our results suggest that in prokaryotes, genome rearrangements may influence functional diversity through the enhanced divergence of boundary genes and the creation of foci for acquiring and deleting species-specific genes.
Key Words: infectious disease • microbial genomics • melioidosis
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Whole-genome comparative analysis for a wide diversity of microbial families has revealed that large-scale genomic alterations are frequently observed between closely related strains and species (Chain et al. 2004
; Yu et al. 2006). Such alterations may contribute to the development of novel virulence (Kuroda et al. 2001; Rowe-Magnus et al. 2003), antibiotic resistance (Kuroda et al. 2001; Rowe-Magnus et al. 2002, 2003; Ito et al. 2003; Paulsen et al. 2003; Gillings et al. 2005), and antigenic variation (Sasaki et al. 2002; Horino et al. 2003) phenotypes and can range from insertions, deletions, and duplications to rearrangement events such as inversions and translocations. Some classic examples include the horizontal transfer of pathogenicity islands containing virulence genes (Lesic and Carniel 2005) and the loss of chromosomal segments harboring virulence-repressing genes and operons (Day et al. 2001). However, whereas large-scale genome gains and losses have long been recognized as important features of microbial evolution, the influence of rearrangement events such as translocations and inversions on functional diversity has been less explored. Specifically, it is unclear how rearrangements, by disrupting existing patterns of gene order and chromosomal organization, might specifically contribute to functional alterations in prokaryotic cellular pathways. Given the high frequency of such chromosomal rearrangements in bacterial families, exploring this issue is likely to be important and may further our understanding of gene–phenotype relationships in microbes.
Intriguingly, recent studies in eukaryotes have suggested the existence of a nonrandom relationship between large-scale genomic rearrangements and fine-scale nucleotide variation (Burt et al. 1999). In certain vertebrate species such as dog, genes localized at rearrangement breakpoints were found to display significantly higher nucleotide substitution rates compared with genes in unaltered genomic regions (Navarro and Barton 2003; Marques-Bonet and Navarro 2005; Webber and Ponting 2005). Similarly, a comparative study between humans and chimpanzees also reported an increased level of gene expression divergence for genes located at genomic breakpoints (Marques-Bonet et al. 2004). Notably, however, the existence of such relationships has not been detected in other eukaryotic studies (Vallender and Lahn 2004; Zhang et al. 2004). We were interested in exploring this potential association in prokaryotes because simpler genomes are likely to provide more specific insights into this phenomenon. In this study, we performed a systematic analysis of chromosomal breakpoints and genome rearrangements in 4 related bacterial species of the Burkholderia genus: Burkholderia cenocepacia (Bc), Burkholderia mallei (Bm), Burkholderia pseudomallei (Bp), and Burkholderia thailandensis (Bt). Of these 4 species, 3 (Bp, Bm, and Bc) are recognized human pathogens, whereas Bt is avirulent but otherwise phenotypically similar to Bp (Wuthiekanun et al. 1996; Kim et al. 2005). By identifying genomic regions of conserved gene order (synteny blocks) across the 4 species, we discovered that Burkholderia genes located at "synteny block" boundaries exhibited a significantly higher degree of molecular divergence compared with genes within blocks. Referring to this relationship as boundary element associated divergence (BEAD), we also show that BEAD is detectable in 2 other prokaryotic families (Pseudomonas and Shigella), suggesting that it may be a common phenomenon. Finally, an analysis of species-specific gene insertions and deletions also found that they were preferentially localized to synteny breakpoints. Our results thus suggest that in prokaryotes, genome rearrangements may influence functional diversity by both the enhanced divergence of boundary genes and by creating focal regions for the acquisition and deletion of species-specific genes.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Microbial Genome Sequences
Genome sequences for Bp K96243, Bm ATCC 23344, and Bt E264 were obtained from GenBank under accession numbers (Chr 1/Chr 2) Bp (BX571965 [GenBank] /BX571966) (Holden et al. 2004
), Bm (CP000010
[GenBank]
/CP000011) (Nierman et al. 2004), and Bt (NC_007651
[GenBank]
/NC_007650) (Yu et al. 2006) (supplementary table S1, Supplementary Material online). The Bc J2315 genome was downloaded from the Burkholderia Cenocepacia Sequencing Web site (ftp://ftp.sanger.ac.uk/pub/pathogens/bc). Permission was obtained from the Sanger Centre to utilize the unpublished Bc genome sequence in this study. Genome sequences for Shigella species Shigella boydii Sb227 (Sbo) (Yang et al. 2005), Shigella dysenteriae Sd197 (Sdy) (Yang et al. 2005), Shigella flexneri 2a strain 2457T (So) (Wei et al. 2003), and Shigella sonnei Ss046 (Ss) (Yang et al. 2005) were obtained from GenBank under accession numbers—Sbo (NC_007613
[GenBank]
), Sdy (CP_000034), So (NC_004741
[GenBank]
), and Ss (NC_007384
[GenBank]
). Genome sequences of Pseudomonas species Pseudomonas aeruginosa PA01 (Pa) (Stover et al. 2000), Pseudomonas fluorescens Pfl-5 (Pfl) (Paulsen et al. 2005), Pseudomonas putida KT2440 (Pp) (Nelson et al. 2002), and Pseudomonas entomophila L48 (Pseen) (Vodovar et al. 2006) were obtained from GenBank under accession numbers—Pa (NC_002516
[GenBank]
), Pfl (NC_004129
[GenBank]
), Pp (NC_002947
[GenBank]
), and Pseen (NC_008027
[GenBank]
).
Identification of Cross-Species Homologs (Metagenes)
Using the Bp genome as a reference, 3,460 Chr 1 and 2,395 Chr 2 open reading frames (ORFs) were queried against the Bc, Bm, and Bt genomes using TBlastN (Altschul et al. 1990). To minimize ambiguous predictions such as ORFs with matches to multiple genomic locations (supplementary table S2, Supplementary Material online), we constrained the resulting matches to have 1) a minimum length of 50 amino acids, 2) a minimal e value cutoff of 1 x 10–6, and 3) a minimum percent identity of 50%. Bc Chr 3 was excluded from analysis as only 27 "metagenes" (<1%) were gained from its inclusion. Homology assignments returned 2,675 genes and these were validated by a reciprocal Blast assay resulting in 2,590 genes. Control analyses using Bc, Bm, or Bt as starting reference genomes yielded similar metagene sets (data not shown). The enrichment of metagenes on Chr 1 was computed using a Fisher's test. To correct for differences in chromosomal size, the Chr 1 and Chr 2 metagenes were scaled to the number of metagenes/Mbp. Similar procedures were used to identify Shigella and Pseudomonas metagenes and genes that were identified as acquired or deleted in the various Burkholderia species. Codeml (Yang 1997) was used to calculate pairwise nucleotide substitution rates (Ks) between metagenes from different species. Substitution rates of >2 were filtered as they were deemed unreliable (Hillier et al. 2004; Zheng et al. 2004).
Chromosomal Rearrangement Analysis
Relative metagene order of the 4 Burkholderia species was analyzed using the Multiple Genome Rearrangement (MGR) program (Bourque and Pevzner 2002). MGR considers inversions, translocations, fusions, and fissions and computes a most parsimonious scenario that best explains the observed gene arrangements. Because "translocated" metagenes only represent 6% of all-metagenes and might be associated with transpositions (a rearrangement event not directly considered by MGR), they were removed from this initial analysis resulting effectively in an inversion-only metric for MGR. In this study, MGR was further constrained to use the tree topology previously published based on the analysis of 16S rRNA genes (Yu et al. 2006) because the high similarities between the Bp, Bt, and Bm gene arrangements led to an underdetermined internal edge. Similar procedures were used to generate synteny maps for the Pseudomonas and Shigella families.
Boundary Element Associated Divergence (BEAD) Analysis
Relationships between metagene divergence and proximity to rearrangement breakpoints were determined by computing the divergence of each metagene located at distances varying from 0 (immediately adjacent) up to 10 metagenes away from a breakpoint. We refer to these as boundary metagenes (Bmets). Divergence values were then averaged across all Bmets and compared against the population of all-metagenes. P values were computed using a standard 2-tailed t-test and a nonparametric Wilcoxon rank sum test. In the Wilcoxon test, the metagenes were ranked in decreasing order of divergence (i.e., the most diverged metagene pair being the top rank) and the distribution of Bmets within this ranking was assessed.
KEGG Pathway Mapping of Metagenes
Bp genes associated with Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways at the lowest level (third level) were retrieved from http://www.genome.jp/kegg/. The pathway-mapped Bp genes were then mapped to the 2,435 metagenes. Each metagene was then compared with its adjacent metagenes to determine if they belonged to the same reference pathway. When a metagene was involved in multiple KEGG pathways, a same metagene pair was considered if any of its reference pathways matched the adjacent metagene.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Chromosomal Mapping of Orthologous Genes Reveals Distinct Rates of Sequence Divergence between Burkholderia Chromosomes
To identify homologous genes across Bp, Bm, Bt, and Bc, we employed a reciprocal Blast strategy (see Materials and Methods). We identified 2,590 genes with unambiguous single orthologs across the 4 species and subsequently refer to these gene entities as metagenes. From this, we estimate that approximately 44% of genes in the Bp genome (2,590/5,855) are highly conserved across all 4 species. Variations in the similarity threshold cutoff for assigning homologs yielded comparable metagene sets (see Materials and Methods) indicating that the 2,590-metagene set is robust. Note that Bp is used as the reference genome in this study because of our prior interest in this organism as the causative agent of melioidosis (Wiersinga et al. 2006
).
The Bp, Bm, and Bt genomes comprise 2 chromosomes, whereas the Bc genome has 3. Because the inclusion of Bc Chr 3 only affected the overall metagene count by 
1%, this chromosome was excluded from subsequent analysis. Studying the distribution of the metagenes in the genome showed an uneven spread across the 2 chromosomes. Specifically, in Bp, the majority of metagenes (75.5%, 1,955/2,590) were found on Chr 1, with the remaining 24.5% on Chr 2. This difference is highly significant even after correcting for chromosomal size differences (P = 1.64 x 10–34; see Materials and Methods). A similar analysis of Bm, Bt, and Bc revealed that 72% of the metagenes were consistently found on Chr 1 in all genomes, whereas 22% were found on Chr 2 (fig. 1). Under a parsimony assumption, this result implies that the chromosomal assignments are likely to be ancestral. In contrast, only 6% of the metagenes (155 genes) appear to have experienced an interchromosomal displacement event. We refer to this set as translocated metagenes and the remaining 2,435 metagenes (94%) as "nontranslocated" metagenes.
|
Interestingly, we found that the depletion of metagenes on Chr 2 could be explained, at least in part, by genes on Chr 2 experiencing an accelerated rate of sequence divergence. This is supported by 2 findings. First, a comparison of pairwise amino acid similarities between Bp metagenes and their orthologs in the 3 other species revealed a significantly lower average amino acid similarity in Chr 2 metagenes compared with Chr 1 metagenes (table 1; pairwise P values in Bc: 4.4 x 10–27, Bm: 2.8 x 10–11, and Bt: 0.031). Second, a similar comparison of nucleotide substitution rates also confirmed significantly increased nucleotide substation rates (Ks) in Chr 2 metagenes compared with Chr 1 (supplementary table S3, Supplementary Material online). This striking interchromosomal difference thus promoted us to analyze the 2 chromosomes separately in subsequent analyses.
|
Functionally Related Genes Are Embedded in Conserved Synteny Blocks
To identify regions of conserved gene order and large-scale rearrangement events across the 4 species, we applied the MGR algorithm (Bourque and Pevzner 2002
) on the set of 2,435 nontranslocated Burkholderia metagenes (see Materials and Methods). Initially, the 2,435 metagenes were condensed into 255 synteny blocks, referring to chromosomal regions of conserved gene order across all 4 species (average # of metagenes/block = 32). The final rearrangement scenario recovered a total of 242 rearrangement events consisting of: 9 inversions (6 on Chr 1 and 3 on Chr 2) between Bt and Bp, 52 inversions between Bm and Bp, and 181 inversions between Bc and Bp (fig. 2A).
|
The Bp genome contains numerous clusters of functionally related genes, some which are likely to correspond to operons (Rodrigues et al. 2006
). We hypothesized that if the cellular functions mediated by these clusters are important units that are conserved and required through evolution, genes in these units should thus be preferentially be preserved within synteny blocks (Tamames 2001). To test this model, we used the KEGG cellular database to functionally annotate the individual metagenes in terms of their cellular and metabolic pathways (Kanehisa and Goto 2000) (see Materials and Methods). We then compared the KEGG pathway assignments for adjacent pairs of metagenes that were either within a single synteny block or between different blocks. For both Chr 1 and Chr 2, we found that metagenes within the same synteny block were far more likely to share related cellular functions compared with metagenes between blocks (Chr 1: 45% vs. 15%, P = 1.32 x 10–07; Chr 2: 55% vs. 6%, P = 1.99 x 10–16; by Binomial test; supplementary table S4 [Supplementary Material online]). This effect remained significant albeit weaker even after accounting for the fact that gaps between synteny blocks are typically larger than those within blocks (data not shown). Overall, these results suggest that synteny blocks tend to represent units of genes with similar functions and that rearrangement events tend to be biased in favor of separating genes with distinct functions. Besides identifying ancestral rearrangement events, the MGR algorithm also allowed a putative Burkholderia median ancestor to be reconstructed (fig. 2A). A comparison with Bp reveals the existence of 5 large synteny blocks in the predicted ancestor. We observed that some of the metagenes in synteny block 4, which were also coadjacent in Bp, Bc, and Bt, are localized to distinct chromosomal locations in Bm (see red line in fig. 2A). Supporting this putative ancestral organization, many of the genes in synteny block 4 were functionally related by being involved in ABC transporter processes (fig. 2B). Thus, although the computational prediction should be treated with caution, this result suggests that this segment of synteny block 4 in the putative ancestor may have experienced a genomic rearrangement in the Bm lineage.
Correlating Rearrangement Breakpoints with Nucleotide Divergence Reveals BEAD
To explore the relationship between large-scale genome rearrangement events and fine-scale genetic divergence, we then identified a population of Bmets, corresponding to metagenes localized adjacent or in close proximity to synteny breakpoints (see Materials and Methods). We calculated levels of sequence divergence in Bmets based on pairwise comparisons between Bp and the other 3 species and contrasted these values against the all-metagene distribution. We discovered that for all 3 comparisons, the Bmet population consistently exhibited enhanced amino acid divergence (fig. 3) and that this was true especially when the Bmet population was confined to metagenes immediately adjacent to a synteny break. The strongest effect was observed between Bp and Bc (Chr 1: 78.93% vs. 83.57%, P = 1.94 x 10–07; Chr 2: 74.76% vs. 78.18%, P = 3.06 x 10–05), followed by the Bt and Bp comparison (Chr 1: 93.74% vs. 95.28%, P = 1.03 x 10–03; Chr 2: 93.01% vs. 93.95%, P = 7.82 x 10–03) (table 2; fig. 3; supplementary fig. S1 [Supplementary Material online]). A similar trend was also observed between Bm and Bp (Chr 1: 99.41% vs. 99.56%, P = 0.249; Chr 2: 99.4% vs. 99.46%, P = 0.5095) where the low significance in the Bm comparison can be explained by the extremely high sequence identity of Bm and Bp metagenes (>99%). Indeed, when we adopted an alternative rank-based statistical approach where the Bm metagenes were ranked in order of their divergence to Bp, this observed difference was highly significant (table 2). Finally, besides amino acid identify, we also compared nucleotide substitution rates between Bmets and the all-metagene population. Similar to the amino acid analysis, Bmets also exhibited higher nucleotide substitution rates compared with all-metagenes, with the strongest difference found in Bc, followed by Bt and Bm (supplementary table S5, Supplementary Material online).
|
|
The observation that Bmets exhibit increased amino acid divergence compared with the global metagene average indicates a potential relationship between the locations of rearrangement breakpoints and fine-scale genetic divergence. We will refer to this phenomenon as BEAD. To investigate if local variations in GC content could explain the occurrence of BEAD, we compared %GC and %GC4D for Bmets and all-metagenes in Bp. However, there was no significant difference in GC content between Bmets and all-metagenes for both Chr 1 and Chr 2 (Chr 1: 67.9% vs. 68.2%, P = 0.36 and Chr 2: 68.9% vs. 68.7%, P = 0.59; supplementary fig. S2 [Supplementary Material online]). This suggests that, in contrast to higher eukaryotes (Webber and Ponting 2005
), mechanisms independent of GC content may underlie BEAD in Burkholderia (see Discussion).
Evidence of BEAD Effects in Translocated Burkholderia Metagenes
To further validate the prevalence of BEAD, we revisited the set of 155 metagenes that had experienced an interchromosomal translocation (or transposition) event in one of the Burkholderia lineages. The 155 translocated metagenes could be classified into 3 distinct groups: A) 69 metagenes translocated to Chr 2 in the Bc lineage, B) 54 metagenes translocated to Chr 1 also in the Bc lineage, and C) 32 metagenes translocated to Chr 2 in the Bm lineage. Similar to Bmets, we tested if these translocated metagenes might be associated with increased divergence rates relative to the entire metagene population. The average identity of group A metagenes was 69.01%, which was significantly lower than both the average identity of nontranslocated metagenes on Bc Chr 1 (83.57%, P = 2.55 x 10–17) and on Bc Chr 2 (78.18%, P = 1.77 x 10–09). Similarly, the average identity of group B metagenes was 76.35%, which was significantly lower than the average identity of nontranslocated metagenes on Bc Chr 1 (83.57%, P = 1.39 x 10–05) and lower than on Bc Chr 2 (78.18%, P = 0.24). In a comparison to Bp, the average amino acid identity of group C metagenes was 99.38%, which was lower than the average identify of nontranslocated metagenes on Bm Chr 1 (99.56%, P = 0.62) and Bm Chr 2 (99.46%, P = 0.43). Similar results were observed when we applied the Wilcoxon rank sum test in all 3 groups (table 3). These results indicate that the translocated metagenes appear to be associated with elevated genetic divergence, an observation that is consistent with BEAD.
|
Conservation of BEAD across Other Prokaryotic Species
Besides Burkholderia, chromosome rearrangements are also frequently found in other bacterial families. To investigate if BEAD could also be observed in other prokaryotes, we analyzed 2 other Gram-negative bacterial families: Pseudomonas and Shigella. For Pseudomonas, we retrieved genome sequences for Pa PA01, Pfl Pfl-5, Pp KT2440, and Pseen L48. Using an approach identical to the one used on Burkholderia, we identified 2,023 Pseudomonas metagenes across the 4 species and condensed them into 577 synteny blocks. Similar to our findings in Burkholderia, we found that Pseudomonas Bmets were associated with significantly elevated sequence divergence compared with the total Pseudomonas metagene population, particularly when the Bmet population was confined to metagenes immediately adjacent to a synteny block boundary (table 4). Specifically, the average sequence identities of Bmets compared with all-metagenes for Pseen, Pp, and Pfl were 74.22% versus 76.27% (P = 0.0182); 72.96% versus 75.17% (P = 0.0299); and 73.23% versus 75.39% (P = 0.0256), respectively (table 3; fig. 4). This result indicates that the BEAD is also observed in Pseudomonas species.
|
|
For Shigella, we retrieved genome sequences for Sbo Sb227, Sdy Sd197, So 2a strain 2457T, and Ss Ss046. Using Sdy as a reference, we identified 2,034 Shigella metagenes across the 4 species and condensed them into 137 synteny blocks. Bmets in this analysis displayed a similar profile (table 3; supplementary fig. S3 [Supplementary Material online]). The decreased significance of the Shigella comparison (Wilcoxon P = 0.0354 to 5.87 x 10–03) compared with Pseudomonas (P = 7.77 x 10–15 to 3.33 x 10–16) is likely a consequence of the 4 Pseudomonas species being far more genetically divergent from one another than the Shigella family (Overall sequence identity is 75% for Pseudomonas compared with 99% for Shigella; table 3). For the Shigella species, this high overall similarity results in more subtle differences between the Shigella Bmets and non-Bmets.
Rearrangement Breakpoints Are Hot spots for Gene Acquisition and Deletion
The availability of a Burkholderia synteny map also provided us with an opportunity to investigate the remaining 56% of Bp genes that were not classified as metagenes. Such "nonmetagenes" might represent genes I) specific to a single lineage, II) present in 2 lineages but missing in the other 2, or III) present in 3 lineages but missing in one. To investigate the possible relationship between synteny blocks and the locations of these species-specific genes, we focused on genes acquired or deleted on a single-specific lineage (Scenarios (I) and (II) above; see table 5). We mapped the locations of the acquired and deleted genes and classified them to be either in gaps within synteny blocks or in gaps between blocks. In Bp, the percentage of acquired/deleted genes located between synteny blocks was higher than expected by chance given the relative sizes of these regions (Chr 1: 67.7% vs. 51.05%, P = 1.05 x 10–09; Chr 2: 90% vs. 81.42%, P = 4.55 x 10–04; see table 5). Similar results were observed for acquired/deleted genes in Bt (Chr 1: 59.48% vs. 51.05%, P = 3.34 x 10–03; Chr 2: 87.83% vs. 81.42%, P = 2.63 x 10–04) and also for genes in Bm (Chr 1: 59.54% vs. 47.54%, P = 1.02 x 10–03; Chr 2: 78.18% vs. 76.38%, P = 0.295). This trend was also observed in Bc Chr 2 (82.19% vs. 78.35%, P = 0.0202) but not for Bc Chr 1 (46.83% vs. 53.51%), potentially because this genome has the highest proportion of ORFs that are not metagenes. Thus, species-specific genes appear to be preferentially acquired and/or deleted at synteny breakpoint regions.
Genome analysis of individual Burkholderia species has identified numerous insertion sequence (IS) elements that may have contributed to the process of gene acquisition, deletion, and rearrangements (Holden et al. 2004; Nierman et al. 2004). We hypothesized that similar to deletions and acquisitions, IS elements might also be found preferentially localized to breakpoints. We manually extracted 43 IS elements from both chromosomes of Bp (22 IS elements on Chr 1 and 21 IS elements on Chr 2) and juxtaposed the location of these IS elements against gaps within synteny blocks or gaps between blocks. We found that the percentage of IS elements between synteny blocks was again higher than expected by chance given the relative sizes of these regions (Chr 1: 81.81% vs. 51.05%, P = 2.93 x 10–03; Chr 2: 95.24% vs. 81.42%, P = 0.0773; by Binomial test; supplementary table S6 [Supplementary Material online]). Our results thus confirm that synteny breakpoints are hot spots for IS elements. These results are supported by related work on Plasmodium showing similar phenomena (Kooij et al. 2005).
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
The study of large-scale chromosomal rearrangement events in prokaryotes has traditionally been complicated by high levels of genome rearrangement and horizontal gene transfer (Ochman et al. 2000
; Boucher et al. 2003) resulting in a low rate of gene order conservation across species (Sankoff 2003). In the case of Burkholderia, the relatively high degree of conservation of gene order, coupled with the presence of multiple chromosomes, makes this microbial family an intriguing model system for studying prokaryotic chromosomal evolution within a single organism. To our knowledge, this is the first study exploring the relationship between large-scale rearrangement events and fine-scale nucleotide substitution in prokaryotes. By comparing Bp, Bm, Bt, and Bc, we identified 2,590 orthologous genes that were conserved across all 4 species. Unexpectedly, these metagenes were unevenly distributed between the 2 Burkholderia chromosomes, with metagenes being significantly more prevalent on Chr 1 than Chr 2. Previous work in higher eukaryotes has shown that homologous chromosomes across different species can often experience different rates of large-scale evolutionary change (Bourque et al. 2005; Fischer et al. 2006). What is less explored, however, is whether distinct chromosomes within the same organism can also display different rates of evolutionary change. In the case of Burkholderia, both chromosomes contain different repertoires of gene content and exhibited varying evolutionary rates. This is consistent with the fact that the Burkholderia chromosomes may be functionally partitioned, with housekeeping genes preferentially localized on Chr 1 thereby requiring strong conservation, and genes for accessory functions on Chr 2 (Holden et al. 2004). Alternatively, it might be that Chr 1 and 2 were acquired at different times by the Burkholderia ancestor. For example, Chr 1 might represent the true ancestral chromosome, whereas Chr 2 might have been acquired more recently—such an event might also contribute to the differences in metagene divergence between the 2 chromosomes.
We generated a 4-way whole-genome synteny map and discovered an association between genome rearrangement breakpoints and increased nucleotide variation; referred to as BEAD. Repeating the analysis in Pseudomonas and Shigella, we then showed that BEAD was detectable in other bacterial families. Although the nonrandom association between large-scale and fine-scale variation had been previously reported for some higher eukaryotic species (see below), our study suggests that BEAD is likely a prevalent process in prokaryotes as well. The BEAD effect in prokaryotes was most significant for metagenes located immediately adjacent to rearrangement breakpoints, and similarly, in eukaryotic genomes elevations in sequence diversity have been shown to decline statistically from the ends of chromosomes, particularly in subtelomeric regions (Navarro and Barton 2003; Hillier et al. 2004; Marques-Bonet et al. 2004; Webber and Ponting 2005). Despite this similarity, it is possible that the mechanisms underlying this phenomenon may be different between eukaryotes and prokaryotes. Specifically, we did not observe in Burkholderia a relationship between elevated GC content and nucleotide divergence, unlike that reported for eukaryotes. In eukaryotes, this bias has been attributed to nucleotide substitution rates in different regions being strongly affected by differences in CpG content due to the hypermutability of methylated cytosines in CpG dinucleotides (Cooper and Youssoufian 1988; Sved and Bird 1990). The mechanisms underlying BEAD in prokaryotes deserve to be further studied but may be linked to the preferential localization of IS elements at breakpoint regions. One possibility is that the introduction of such elements may have affected the local fidelity of the DNA replication and recombination machinery, ultimately resulting in an enhanced level of sequence divergence between genomes. An alternative possibility is that the high recombination rate of IS elements might also directly induce genome rearrangements due to de novo insertions and recombination across these elements. This latter possibility may be especially relevant to Bm, which contains many more IS elements than the other Burkholderia species.
An interesting logical cause-and-effect question following from this result: 1) does genomic breakage precede the appearance of high nucleotide divergence, 2) does nucleotide divergence at a genomic loci precede a chromosomal break, or 3) do these 2 processes act in an independent but correlated fashion? For example, it is possible that syntenic breaks may preferentially occur at genomic loci where mutations are intrinsically better tolerated, due to a lack of essential genes. Supporting this, the Bmets were significantly underrepresented in genes involved in essential functions such as nucleotide metabolism (P = 0.009) and protein translation (P = 0.03) (see supplementary table S7, Supplementary Material online). However, although this finding supports the notion that Bmets should be intrinsically more tolerant to nucleotide variation, it still does not unambiguously resolve whether genomic breakage precedes nucleotide divergence, or vice versa. Clearly, distinguishing clearly between these different models will be challenging. In a preliminary attempt to address this issue, we partitioned the Bmets identified in the 4-way comparison into 2 groups: Bmets exclusive to rearrangement events occurring on the path between Bp and Bc and Bmets "created" due to events on the Bt or Bm lineages (see supplementary fig. S4, Supplementary Material online). Interestingly, in a Bp versus Bc comparison, the percent identity of Bmets created on the Bp–Bc path was significantly lower than the Bmets associated with the non-Bc lineages (i.e., Bp–Bm and Bp–Bt, see supplementary table S8 [Supplementary Material online]). Similar results were also obtained in a Bp–Bt centric analysis (see supplementary table S9, Supplementary Material online). This result may suggest that nucleotide divergence does not precede genomic breakage. Unfortunately, the current data set did not allow us to distinguish between the remaining 2 scenarios because of the limited number of rearrangements on the internal edge for this family.
In conclusion, we have in this study demonstrated in prokaryotes the existence of a link between large-scale genome rearrangements and fine-scale nucleotide variation localized to rearrangement breakpoints. As more bacterial genomes are sequenced, we will be able to further assess the general applicability of BEAD and its potential role in microbial evolution.
|
Supplementary Material |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Supplementary tables S1–S9 and figures S1–S4 are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).
|
|
Acknowledgements |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
This work is supported by a block grant to the Genome Institute of Singapore from the Agency for Science, Technology, and Research of Singapore. We thank Dr. Julian Parkhill at the Sanger Centre, United Kingdom for sharing unpublished Bc genome information.
|
Footnotes |
|---|
Jennifer Wernegreen, Associate Editor
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Mol Biol (1990) 215:403–410.[CrossRef][Web of Science][Medline]
Boucher Y, Douady CJ, Papke RT, Walsh DA, Boudreau ME, Nesbo CL, Case RJ, Doolittle WF. Lateral gene transfer and the origins of prokaryotic groups. Annu Rev Genet (2003) 37:283–328.[CrossRef][Web of Science][Medline]
Bourque G, Pevzner PA. Genome-scale evolution: reconstructing gene orders in the ancestral species. Genome Res (2002) 12:26–36.
Bourque G, Zdobnov EM, Bork P, Pevzner PA, Tesler G. Comparative architectures of mammalian and chicken genomes reveal highly variable rates of genomic rearrangements across different lineages. Genome Res (2005) 15:98–110.
Burt DW, Bruley C, Dunn IC, et al. The dynamics of chromosome evolution in birds and mammals. Nature (1999) 402:411–413. (13 co-authors).[CrossRef]
Chain PS, Carniel E, Larimer FW, et al. Insights into the evolution of Yersinia pestis through whole-genome comparison with Yersinia pseudotuberculosis. Proc Natl Acad Sci USA (2004) 101:13826–13831. (23 co-authors).
Cooper DN, Youssoufian H. The CpG dinucleotide and human genetic disease. Hum Genet (1988) 78:151–155.[CrossRef][Web of Science][Medline]
Day WA Jr, Fernandez RE, Maurelli AT. Pathoadaptive mutations that enhance virulence: genetic organization of the cadA regions of Shigella spp. Infect Immun (2001) 69:7471–7480.
Fischer G, Rocha EP, Brunet F, Vergassola M, Dujon B. Highly variable rates of genome rearrangements between hemiascomycetous yeast lineages. PLoS Genet (2006) 2:e32.[CrossRef][Medline]
Gillings MR, Holley MP, Stokes HW, Holmes AJ. Integrons in Xanthomonas: a source of species genome diversity. Proc Natl Acad Sci USA (2005) 102:4419–4424.
Hillier LWW, Miller E, Birney, et al. Sequence and comparative analysis of the chicken genome provide unique perspectives on vertebrate evolution. Nature (2004) 432:695–716. (175 co-authors).[CrossRef][Medline]
Holden MT, Titball RW, Peacock SJ, et al. Genomic plasticity of the causative agent of melioidosis, Burkholderia pseudomallei. Proc Natl Acad Sci USA (2004) 101:14240–14245. (48 co-authors).
Horino A, Sasaki Y, Sasaki T, Kenri T. Multiple promoter inversions generate surface antigenic variation in Mycoplasma penetrans. J Bacteriol (2003) 185:231–242.
Ito T, Okuma K, Ma XX, Yuzawa H, Hiramatsu K. Insights on antibiotic resistance of Staphylococcus aureus from its whole genome: genomic island SCC. Drug Resist Updat (2003) 6:41–52.[CrossRef][Web of Science][Medline]
Kanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res (2000) 28:27–30.
Kim HS, Schell MA, Yu Y, Ulrich RL, Sarria SH, Nierman WC, DeShazer D. Bacterial genome adaptation to niches: divergence of the potential virulence genes in three Burkholderia species of different survival strategies. BMC Genomics (2005) 6:174.[CrossRef][Medline]
Kooij TW, Carlton JM, Bidwell SL, Hall N, Ramesar J, Janse CJ, Waters AP. A Plasmodium whole-genome synteny map: indels and synteny breakpoints as foci for species-specific genes. PLoS Pathog (2005) 1:e44.[CrossRef][Medline]
Kuroda M, Ohta T, Uchiyama I, et al. Whole genome sequencing of meticillin-resistant Staphylococcus aureus. Lancet (2001) 357:1225–1240. (37 co-authors).[CrossRef][Web of Science][Medline]
Lesic B, Carniel E. Horizontal transfer of the high-pathogenicity island of Yersinia pseudotuberculosis. J Bacteriol (2005) 187:3352–3358.
Marques-Bonet T, Caceres M, Bertranpetit J, Preuss TM, Thomas JW, Navarro A. Chromosomal rearrangements and the genomic distribution of gene-expression divergence in humans and chimpanzees. Trends Genet (2004) 20:524–529.[CrossRef][Web of Science][Medline]
Marques-Bonet T, Navarro A. Chromosomal rearrangements are associated with higher rates of molecular evolution in mammals. Gene (2005) 353:147–154.[CrossRef][Medline]
Navarro A, Barton NH. Chromosomal speciation and molecular divergence–accelerated evolution in rearranged chromosomes. Science (2003) 300:321–324.
Nelson KE, Weinel C, Paulsen IT, et al. Complete genome sequence and comparative analysis of the metabolically versatile Pseudomonas putida KT2440. Environ Microbiol (2002) 4:799–808. (43 co-authors).[CrossRef][Medline]
Nierman WC, DeShazer D, Kim HS, et al. Structural flexibility in the Burkholderia mallei genome. Proc Natl Acad Sci USA (2004) 101:14246–14251. (33 co-authors).
Ochman H, Lawrence JG, Groisman EA. Lateral gene transfer and the nature of bacterial innovation. Nature (2000) 405:299–304.[CrossRef][Medline]
Paulsen IT, Banerjei L, Myers GS, et al. Role of mobile DNA in the evolution of vancomycin-resistant Enterococcus faecalis. Science (2003) 299:2071–2074. (32 co-authors).
Paulsen IT, Press CM, Ravel J, et al. Complete genome sequence of the plant commensal Pseudomonas fluorescens Pf-5. Nat Biotechnol (2005) 23:873–878. (29 co-authors).[CrossRef][Web of Science][Medline]
Rodrigues F, Sarkar-Tyson M, Harding SV, et al. Global map of growth-regulated gene expression in Burkholderia pseudomallei, the causative agent of melioidosis. J Bacteriol (2006) 188:8178–8188. (14 co-authors).
Rowe-Magnus DA, Guerout AM, Biskri L, Bouige P, Mazel D. Comparative analysis of superintegrons: engineering extensive genetic diversity in the Vibrionaceae. Genome Res (2003) 13:428–442.
Rowe-Magnus DA, Guerout AM, Mazel D. Bacterial resistance evolution by recruitment of super-integron gene cassettes. Mol Microbiol (2002) 43:1657–1669.[CrossRef][Web of Science][Medline]
Sankoff D. Rearrangements and chromosomal evolution. Curr Opin Genet Dev (2003) 13:583–587.[CrossRef][Web of Science][Medline]
Sasaki Y, Ishikawa J, Yamashita A, et al. The complete genomic sequence of Mycoplasma penetrans, an intracellular bacterial pathogen in humans. Nucleic Acids Res (2002) 30:5293–5300. (11 co-authors).
Stover CK, Pham XQ, Erwin AL, et al. Complete genome sequence of Pseudomonas aeruginosa PA01, an opportunistic pathogen. Nature (2000) 406:959–964. (31 co-authors).[CrossRef][Medline]
Sved J, Bird A. The expected equilibrium of the CpG dinucleotide in vertebrate genomes under a mutation model. Proc Natl Acad Sci USA (1990) 87:4692–4696.
Tamames J. Evolution of gene order conservation in prokaryotes. Genome Biol (2001) 2:1–11.[Medline]
Vallender EJ, Lahn BT. Effects of chromosomal rearrangements on human-chimpanzee molecular evolution. Genomics (2004) 84:757–761.[CrossRef][Web of Science][Medline]
Vodovar N, Vallenet D, Cruveiller S, et al. Complete genome sequence of the entomopathogenic and metabolically versatile soil bacterium Pseudomonas entomophila. Nat Biotechnol (2006) 24:673–679. (16 co-authors).[CrossRef][Medline]
Webber C, Ponting CP. Hotspots of mutation and breakage in dog and human chromosomes. Genome Res (2005) 15:1787–1797.
Wei J, Goldberg MB, Burland V, et al. Complete genome sequence and comparative genomics of Shigella flexneri serotype 2a strain 2457T. Infect Immun (2003) 71:2775–2786. (17 co-authors).
Wiersinga WJ, van der Poll T, White NJ, Day NP, Peacock SJ. Melioidosis: insights into the pathogenicity of Burkholderia pseudomallei. Nat Rev Microbiol (2006) 4:272–282.[CrossRef][Web of Science][Medline]
Wuthiekanun V, Smith MD, Dance DA, Walsh AL, Pitt TL, White NJ. Biochemical characteristics of clinical and environmental isolates of Burkholderia pseudomallei. J Med Microbiol (1996) 45:408–412.
Yang F, Yang J, Zhang X, et al. Genome dynamics and diversity of Shigella species, the etiologic agents of bacillary dysentery. Nucleic Acids Res (2005) 33:6445–6458. (27 co-authors).
Yang Z. PAML: a program package for phylogenetic analysis by maximum likelihood. Comput Appl Biosci (1997) 13:555–556.
Yu Y, Kim HS, Chua HH, et al. Genomic patterns of pathogen evolution revealed by comparison of Burkholderia pseudomallei, the causative agent of melioidosis, to avirulent Burkholderia thailandensis. BMC Microbiol (2006) 6:46. (12 co-authors).[CrossRef][Medline]
Zhang J, Wang X, Podlaha O. Testing the chromosomal speciation hypothesis for humans and chimpanzees. Genome Res (2004) 14:845–851.
Zheng Y, Roberts RJ, Kasif S. Identification of genes with fast-evolving regions in microbial genomes. Nucleic Acids Res (2004) 32:6347–6357.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. A. Ragan and R. G. Beiko Lateral genetic transfer: open issues. Phil Trans R Soc B, August 12, 2009; 364(1527): 2241 - 2251. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. A. Duerkop, J. Varga, J. R. Chandler, S. B. Peterson, J. P. Herman, M. E. A. Churchill, M. R. Parsek, W. C. Nierman, and E. P. Greenberg Quorum-Sensing Control of Antibiotic Synthesis in Burkholderia thailandensis J. Bacteriol., June 15, 2009; 191(12): 3909 - 3918. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Shanks, M. N. Burtnick, P. J. Brett, D. M. Waag, K. B. Spurgers, W. J. Ribot, M. A. Schell, R. G. Panchal, F. C. Gherardini, K. D. Wilkinson, et al. Burkholderia mallei tssM Encodes a Putative Deubiquitinase That Is Secreted and Expressed inside Infected RAW 264.7 Murine Macrophages Infect. Immun., April 1, 2009; 77(4): 1636 - 1648. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||