Figures
Abstract
Burkholderia pseudomallei is the causative agent of melioidosis and a potential bioterrorism agent. In the development of medical countermeasures against B. pseudomallei infection, the US Food and Drug Administration (FDA) animal Rule recommends using well-characterized strains in animal challenge studies. In this study, whole genome sequence data were generated for 6 B. pseudomallei isolates previously identified as candidates for animal challenge studies; an additional 5 isolates were sequenced that were associated with human inhalational melioidosis. A core genome single nucleotide polymorphism (SNP) phylogeny inferred from a concatenated SNP alignment from the 11 isolates sequenced in this study and a diverse global collection of isolates demonstrated the diversity of the proposed Animal Rule isolates. To understand the genomic composition of each isolate, a large-scale blast score ratio (LS-BSR) analysis was performed on the entire pan-genome; this demonstrated the variable composition of genes across the panel and also helped to identify genes unique to individual isolates. In addition, a set of ~550 genes associated with pathogenesis in B. pseudomallei were screened against the 11 sequenced genomes with LS-BSR. Differential gene distribution for 54 virulence-associated genes was observed between genomes and three of these genes were correlated with differential virulence observed in animal challenge studies using BALB/c mice. Differentially conserved genes and SNPs associated with disease severity were identified and could be the basis for future studies investigating the pathogenesis of B. pseudomallei. Overall, the genetic characterization of the 11 proposed Animal Rule isolates provides context for future studies involving B. pseudomallei pathogenesis, differential virulence, and efficacy to therapeutics.
Citation: Sahl JW, Allender CJ, Colman RE, Califf KJ, Schupp JM, Currie BJ, et al. (2015) Genomic Characterization of Burkholderia pseudomallei Isolates Selected for Medical Countermeasures Testing: Comparative Genomics Associated with Differential Virulence. PLoS ONE 10(3): e0121052. https://doi.org/10.1371/journal.pone.0121052
Academic Editor: R. Mark Wooten, University of Toledo School of Medicine, UNITED STATES
Received: September 15, 2014; Accepted: January 27, 2015; Published: March 24, 2015
Copyright: © 2015 Sahl et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data, including accession numbers are within the paper and its Supporting Information files.
Funding: This project was funded in part with Federal funds from Biomedical Advanced Research and Development Authority (BARDA), Department of Health and Human Services, under Task Order No. HHSO10033001T, contract HHSO100201100005I.Additional support was provided by this project was funded by National Institutes of Health-National Institute of Allergy and Infectious Diseases Grant U54 AI-065359 and U01 AI-075568. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Burkholderia pseudomallei is a pathogen endemic to Southeast Asia and Northern Australia but is increasingly found in other parts of the world including India, South America, and Africa, where it is naturally found in soil and water [1]. The bacterium is the causative agent of melioidosis [2–5], a potentially fatal disease in humans. B. pseudomallei is also considered to be a Tier 1 biothreat agent due to its ease of attainment, ability to cause lethal disease, intrinsic antibiotic resistance [6], and lack of a melioidosis vaccine [7]. The development of appropriate medical countermeasures against melioidosis has been hampered by access to human patients for clinical trials with compounds that are not currently approved for the treatment of melioidosis. To address this concern, the US Food and Drug Administration (FDA) has instituted the “Animal Rule” 21 CFR that calls for well-characterized strains to be used in animal challenge studies [8], including BALB/c mice, which have shown to represent acute human melioidosis [9]. Based on several selection criteria, a recent study selected a panel of six B. pseudomallei strains that would be appropriate for challenge studies under the FDA Animal Rule [7].
In the current study, we used whole-genome sequencing (WGS) to genetically characterize a panel of B. pseudomallei strains to be used as challenge material in therapeutic efficacy studies under the Animal Rule. In addition, we sequenced 5 B. pseudomallei strains associated with inhalational disease for evaluation as potential challenge strains. The purpose of WGS on these isolates was to (1) characterize the genomic background in each isolate; (2) identify the phylogenetic diversity of panel isolates in the context of a global set of genomes and; (3) identify the distribution of characterized virulence factors for correlation with virulence data obtained in animal challenge studies.
Methods
Strain selection
Eleven diverse isolates were selected for sequencing (Table 1). Six of these isolates were previously selected as part of a proposed B. pseudomallei strain panel, based on several selection criteria [7]. For five of these isolates, there are finished genome assemblies available in public databases [10]; these genomes were sequenced to identify any mutations compared to the published genomes. The genome for an additional isolate, NCTC 13392, has previously been published [11]. An additional 5 isolates were selected based on recent isolation and suspected inhalational disease and were associated with acute pneumonia sepsis.
Animal challenge studies
285 BALB/c mice (100% female) were purchased from Charles River Laboratories and were randomly selected and placed into challenge groups (n = 7) based on different isolates and dosing. Mice here housed in Innovive IVC mouse racks using disposable caging (7 mice per cage). Sedated mice were challenged by intranasal inoculation (15 μl per nare) of target doses diluted in Dulbecco’s Phosphate-Buffered Saline (PBS); mice were anesthetized intraperitoneally with ketamine (50–120 mg/kg) and xylazine (5–10 mg/kg). Prior to challenge, cultures were grown for 22 hours shaking at 37°C at 250xRPM; no mice were mock-treated in this study. The culture was then centrifuged and re-suspended in PBS containing 0.01% gelatin. The concentration of each challenge dilution was determined by spread plate enumeration.
Following challenge, mice were monitored every 8 hours between days 1 and 7, then twice daily between days 8 and 21; sample HBPUB10303a was only challenged for 14 days due to unforeseen delays in starting the experiment. Observations were made for clinical signs of illness, including respiratory distress, loss of appetite and activity, and seizures; any animal judged to be moribund by a trained animal technician was humanely euthanized. All study survivors were humanely euthanized with CO2 inhalation on Study Day 21. Kaplan-Meier survival curves were created using the ‘survival’ package in R [12]. Animal challenge studies were conducted at the Battelle Biomedical Research Center (BBRC). All animal work was approved by Battelle’s IACUC prior to study initiation.
DNA extraction, library creation, sequencing
DNA library constructions were performed using the KAPA Library Preparation Kits with Standard PCR Library Amplification/Illumina series (KAPA biosystems, Boston MA, code KK8201). Quality and quantity of genomic DNA were evaluated by agarose gel analysis. One to two micrograms of DNA per sample were fragmented using a SonicMan (Matrical) with following parameters: 75.0 seconds pre chill, 16 cycles, 10.0 sec sonication, 100% power, 75.0 sec lid chill, 10.0 sec plate chill, and 75.0 sec post chill. The fragmented DNA was purified using QIAGEN QIAquick PCR purification columns (QIAGEN, cat. no. 28104) and eluted into 42.5 μl of Elution Buffer. The adapter ligation used 1.5 μl of the 40 μM adapter oligo mix [13]. Only one post-ligation bead cleanup was done. All purification steps were done with the 1.8x SPRI bead protocol in the KAPA protocol. Size selection of fragments was gel based; 30 μl of clean ligated material was run onto a 2% agarose gel. Several gel slices, corresponding to different average DNA fragment sizes (300, 600, and 1000bp fragments) were extracted from the gel and purified with a QIAGEN Gel Extraction kit (QIAGEN, cat. no. 28704) and eluted in 30 μl of Elution Buffer. Due to the high GC content of the samples, the PCR was optimized to improve yield and genomic coverage. Two microliters of DNA, 2 μl of 10 μM of both primers, 25 μl of NEBNext High-Fidelity 2X PCR Master Mix (New England Biolabs, Ipswich, MA, cat. no. M0541S), and 22 μl of 5 M Betaine (Sigma-Aldrich, St. Louis, MO, cat. no. B0300-1VL) were combined. The following PCR parameters were used: initial denaturation of 2 min at 98°C, 12 cycles of 30 sec at 98°C, 20 sec at 65°C, 30 sec at 72°C, with a final extension of 5 min at 72°C.
Genome assembly
For strains that have been sequenced previously, a comparative assembly approach was employed. Reads were assembled against the reference genome (S1 Table) with AMOScmp [14]. Assembled contigs were then aligned against the reference genome with ABACAS [15] to obtain a genomic scaffold. Gaps in scaffolds were filled with IMAGE [16], which also splits un-filled scaffolds into contigs. In addition to the comparative assembly, reads were also assembled with Abyss v. 1.3.4 [17]. The two assemblies were aligned with Mugsy [18] and regions specific to the de novo assembly were parsed from the MAF file [19], as has been done previously [20]. Putative unique regions in the de novo assembly were aligned against the comparative assembly with BLASTN [21]. Regions that significantly aligned (>90% ID, >90% query length) to the comparative assembly were filtered from the analysis. Remaining regions were combined with the comparative assembly. Assembly errors were corrected from this concatenated assembly with iCORN [22], using ten iterations. For strains that had not been sequenced previously, genomes were assembled de novo with Abyss v 1.3.4 and assembly errors were corrected with iCORN. Assembly details are shown in S1 Table.
In silico multi-locus sequence typing (isMLST)
BLASTN [21] was used to extract sequences from the seven loci in the B. pseudomallei MLST scheme [23] from all genome assemblies. To be considered a match, the alignment from the query genome must match a reference allele 100%. Sequence types were assigned to genomes when exact profile matches were identified. The isMLST functionality was performed with a custom Python script (https://gist.github.com/jasonsahl/33b0d9a8e3ac035bb92c). MLST typing information is shown in S1 Table.
Single nucleotide polymorphism (SNP) and indel identification and annotation
For re-sequencing efforts (Table 1), raw reads were mapped to the finished genome with BWA-MEM v0.7.5 [24]. SNPs and indels were then called with the UnifiedGenotyper in GATK v. 2.7 [25]; nucmer [26] was used to find duplicate regions in the reference genome and any SNPs falling within duplicate regions were filtered from the analysis. For a SNP or indel to be called, we required a minimum coverage of 6x and a minimum proportion threshold of 0.90. Nucleotide variants were annotated with snpEFF [27]. All variants were visually confirmed from BAM files with Tablet [28].
Synteny between previously sequenced genomes
In addition to identifying variants between finished genomes and re-sequencing projects, genome assemblies were aligned to completed genomes with MUMmer [29] and dot plots were visualized with mummerplot to identify any structural variation.
Core genome SNP phylogeny
To visualize the phylogenetic diversity of genomes sequenced in this study, a core genome phylogenetic approach was employed; core regions are defined as sequence conserved in all examined genomes. A diverse set of finished and draft genomes was compiled (S2 Table). Raw reads were mapped to B. pseudomallei K96243 [30] with BWA-MEM [24]. SNPs were called from each BAM file with GATK, using the EMIT_ALL_CONFIDENT_SITES method, with a minimum coverage of 6x and a minimum proportion of 0.90. For genomic assemblies, SNPs were identified from nucmer alignments. Positions in K96243 were directly mapped to the corresponding position in each query genome assembly. A matrix was generated (S1 Dataset) with NASP (http://tgennorth.github.io/NASP/) from all reference positions called and polymorphic sites were identified. SNPs that could not be called by GATK, or failed to pass the depth or proportion filters, were filtered from the matrix, as well as SNPs that fell within identified duplications. The remaining dataset consisted of 62,663 SNPs, 50,290 of them being informative. A maximum likelihood phylogeny was inferred on this dataset with RAxML v8.0.17 [31, 32] using the ASC_GTRGAMMA model and 100 bootstrap replicates. The retention index (RI) value [33] was calculated with Phangorn [34].
SNP and homoplasy density
To identify the conservation of the reference chromosomes, as well as to potentially identify any lateral gene transfer events that may confound the phylogeny, a SNP density (SD) and homoplasy density (HD) approach was employed. The SNP matrix was parsed over 1-kb non-overlapping windows of each chromosome and the number of informative SNPs was then calculated. The dataset was then processed with Paup v4.0b10 [35] to calculate the retention index (RI) value for each SNP. An RI value < 0.5 was considered to be homoplasious and the number of homoplasious SNPs over the same 1-kb window was then calculated. The HD value for each 1-kb window was calculated by dividing the number of homoplasious SNPs by the total number of informative SNPs. The distribution of SD and HD across the two chromosomes in K96243 was visualized with Circos [36].
In silico gene screen
A set of previously described virulence factors [1, 30, 37–42] characterized in B. pseudomallei were compiled (S3 Table). Genes were screened against the genomes sequenced in this study with a large-scale blast score ratio (LS-BSR) approach [43]. Genes were translated with BioPython (www.biopython.org) and aligned against its nucleotide sequence with TBLASTN in order to obtain the maximum alignment (reference) bit score. Each gene was then aligned against each genome with TBLASTN in order to obtain the query alignment bit score. The BSR [44] was obtained by dividing the reference bit score by the query bit score. Genes with a BSR value > 0.90 or < 0.80 in all genomes were removed from the analysis; the complete LS-BSR matrix is available as S2 Dataset. The genes were then correlated with the tree to identify phylogenetic patterns of gene presence/absence.
Genotype and phenotype correlations
Two approaches were performed to determine if there were correlations between genomic information and survival information obtained from animal challenge studies. The survival data were split into three categories: low virulence (100% mouse survival after 21 days), intermediate virulence (<100%, >0% survival after 21 days), and high virulence (0% mouse survival after 21 days). LS-BSR values across all genomes were multiplied by 100 in order to convert all float values to integers. The adjusted LS-BSR values were then correlated with the categorical virulence data using a Kruskal-Wallis test [45] implemented in QIIME v. 1.8.0 [46]. Core genome SNP data were also correlated to categorical data with a chi-square test implemented in SciPy. P-values were corrected with the Benjamini-Hochberg correction [47]. To test for false positives, genomes were randomly assigned to two groups of equal size and the average number of SNPs unique to each group was calculated over 10 iterations.
Unique genomic regions
In addition to screening characterized virulence genes in assembled genomes, a de novo approach was also performed. All coding regions (CDSs) from all genomes in the phylogeny were compared with LS-BSR. Regions were determined to be unique to a given genome if they contained a BSR < 0.4 in all non-targeted genomes. Each unique CDS was then aligned against the GenBank [48] nucleotide database with BLASTN, and the closest hit, based on highest bit score, was identified.
Ethics Statement
The animal protocol (2934–100007643) was approved by the Battelle Institutional Animal Care and Use Committee. The research was conducted in compliance with the Animal Welfare Act and followed the principles in the Guide for the Care and Use of Laboratory Animals from the National Research Council, Office of Laboratory Animal Welfare (OLAW), and USDA. Additionally, the research was conducted following an Institutional Animal Care and Use Committee (IACUC) approved protocol. The institution where the research was conducted is fully accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care International (AAALAC).
Results
Comparisons of re-sequenced isolates with finished genomes
Five of the genomes sequenced in this study represent re-sequencing projects of finished genomes available in public databases (S1 Table). However, due to standard laboratory passages, new nucleotide variants can accumulate [49], and were identified in the current study using raw read data. The results demonstrate that many re-sequenced isolates show little mutation since the genomes were published (Table 2). However, the version of K96243 that was sequenced in the current study showed numerous variant positions (33) compared to the completed genome (Table 2), including the loss of two annotated stop codons. Some of these differences could be errors in the original genome sequence, which we are unable to verify. In addition to the analysis of nucleotide variants, the synteny of genomes was visualized as dot plots (S1 Fig) and demonstrated high synteny between all re-sequenced genome assemblies and finished genomes.
Core genome single nucleotide polymorphism (SNP) phylogeny
To phylogenetically characterize the isolates sequenced in this study, a maximum likelihood phylogeny was inferred from ~63,000 core genome SNPs (Fig. 1) identified from 44 genomes. The results demonstrate that the isolates sequenced in the current study show a broad phylogenetic history compared to previously sequenced isolates. By including phylogenetically diverse isolates in the isolate panel, local patterns of gene distribution do not bias the analysis. The retention index (RI) value of the data and maximum likelihood phylogeny demonstrated signs of homoplasy (RI = 0.62). Recombination in B. pseudomallei has been previously described [23] and homoplasy was anticipated due the recombinatorial nature of the species.
The tree was inferred with RAxML v8 [31, 32] using the ASC_GTRGAMMA model and 100 bootstrap replicates. Filled circles are placed at nodes where the bootstrap support values are >90%.
SNP and homoplasy density
The RI value of the phylogeny demonstrated the presence of homoplasy. Based on this dataset, the presence of homoplasy across the reference genome, K96243, was investigated with a SNP and homoplasy density approach. The results demonstrate that with the isolates tested, chromosome 1 of B. pseudomallei K96243 is more highly conserved than chromosome 2 (Fig. 2). Additionally, the homoplasy is distributed across both chromosomes, with no clear regions associated with specific recombination or lateral gene transfer events.
The outer ring represents the number of informative SNPs across 1-kb genomic intervals. The inner ring indicates the number of homoplasious SNPs, as determined by a retention index (RI) value <0.5 calculated by Paup [35], divided by the total number of informative SNPs over the same 1-kb genomic interval. HD and SD values were visualized with Circos [36].
Unique coding sequences (CDSs)
B. pseudomallei has a highly plastic genome and has the ability to acquire new genes horizontally from other microorganisms, especially as the pathogen persists in the environment. A large-scale blast score ratio (LS-BSR) analysis was performed on the 44 B. pseudomallei genomes in the phylogeny (Fig. 1) to identify any unique CDSs in the 11 isolates sequenced in the current study; the criteria for a CDS to be considered unique is that it must have a BSR value < 0.4 in all non-targeted genomes. A list of closest BLAST hits to unique CDSs not associated with either B. pseudomallei or B. mallei, based on the highest bit score, is shown in Table 3. These regions are likely associated with genomic islands horizontally transferred from related organisms [50].
Virulence gene profile
A comprehensive set of virulence-associated genes (S3 Table) was screened against the 11 genomes sequenced in this study with LS-BSR. To only compare differentially conserved regions, genes were filtered if they had a BSR value > 0.90 in all 11 genomes. The resulting variable set of genes (n = 54) was correlated to the phylogeny and LS-BSR values were visualized as a heatmap (Fig. 3). The results demonstrate that phylogenetically-distinct isolates contain a variable composition of virulence-associated genes.
A maximum likelihood phylogeny was inferred on a concatenation of single nucleotide polymorphisms (SNPs) and was correlated to the heatmap.
Every B. pseudomallei isolate in this study contained the B. pseudomallei bimA (BimABp) allele [51], except B. pseudomallei MSHR668, which contained the alternative B. mallei-type (BimABm). The most severe clinical presentations have been associated with the co-occurrence of BimABm with another virulence-associated gene, filamentous hemagglutinin fhaB3 (BPSS2053 in B. pseudomallei K96243), which is linked with adhesion and heightened virulence [52, 53]. While B. pseudomallei MSHR668 is missing fhaB3, it does contain another fhaB gene (similar to fhaB1 from B. pseudomallei MSHR305 [54]). fhaB3 was observed in all Asian isolates in this study, which is consistent with previous work [54, 55]. Isolates sequenced in this study either contained the Yersinia-like fimbriae cluster (YLF) or the B. thailandensis-like flagellum and chemotaxis (BTFC) gene cluster. These genes were included in our analysis because they are suggested as being active during melioidosis.
Two isolates in this study, 1026b and MSHR305, exhibited reduced sequence homology to the T6SS-1 gene, BPSS1511. The T6SS-1 representative sequence, icmF gene (BPSS1511), which is required for intracellular growth of many pathogens associated with eukaryotic cells [56], showed homology, but lower sequence identity, in 1026b and MSHR305. Four isolates (MSHR5855, MSHR305, 1106a, and HBPUB10134a) exhibited reduced sequence homology for BPSS1493, a hypothetical protein associated with type VI secretion.
Animal challenge studies
To identify differential virulence between ten of the eleven isolates sequenced in this study, BALB/c mice (seven per group) were challenged at different concentrations of inoculum (Table 4). At an average of ~10 colony forming units (CFUs) per group, four of the ten isolates killed all of the mice in the group, 5 of the isolates killed an intermediate number of mice, and one isolate (1106a) killed none of the mice (Table 4, S2 Fig, S4 Table); HBPUB10303a was treated as intermediate in terms of virulence, despite the fact that the isolate was challenged for only 14 days instead of 21 in this experiment. At a high concentration of inoculum (~12,000 CFUs), none of the mice survived when challenged with any of the ten panel isolates. This demonstrates that all of the isolates are virulent by intranasal inoculation, but there is a dose-dependent virulence response.
Genotype and phenotype correlations
Differences were observed in both the virulence gene profile and the animal challenge studies. To identify if any CDSs were associated with differential virulence, a combined LS-BSR/QIIME analysis was performed. A Kruskal-Wallis test [45] demonstrated that numerous CDSs were significantly (false detection rate adjusted (FDR) p<0.05) differentially conserved between groups (Table 5); three of these CDSs (BPSS0771, BPSS1185, BPSS1269) have previously been associated with virulence (Table 5). Additionally, an association was made between core genome SNPs and differential virulence. Forty SNPs were only identified in high virulence isolates (Table 6), which could be due to descent and subsequent loss by intermediate and low virulence isolates, but may also be associated with convergent evolution and virulence (Fig. 3). By randomly assigning genomes to high and low virulence groups, an average of 31 correlated SNPs were identified over ten iterations. This demonstrates that with small sample sets, identified correlations would definitely need to be corroborated with functional characterization.
Discussion
Burkholderia pseudomallei is an important pathogen as both the causative agent of melioidosis and as a potential biothreat agent. In the development of medical countermeasures against melioidosis, a panel of clinically relevant isolates have been identified [7] for challenge studies under the FDA Animal Rule [8]. In this study, we sequenced all 6 of these isolates as well as 5 additional isolates associated with human inhalational melioidosis. A comparative genomics approach was employed to understand the genetic composition of each genome and the distribution of genetic elements between genomes. These results were correlated with animal survival data to determine if phenotype/genotype correlations could be identified.
Ten of the 11 isolates were passed through a BALB/c mouse model in groups of seven mice per isolate. Differential virulence was observed between isolates, with MSHR668 demonstrating the highest virulence (S2 Fig, Table 2), based on time to death. An attempt was made to correlate both the distribution of coding sequences (CDSs), based on large-scale blast score ratio (LS-BSR) values, and single nucleotide polymorphisms (SNPs), with differential virulence. Three CDSs previously associated with virulence were differentially conserved between disease severity groups (Table 4). Additionally, SNPs were identified that were only present in high-virulence isolates (Table 6). While the limited number of isolates tested in this study precludes definitive correlations between genotype and phenotype, differentially conserved CDSs and/or SNPs may inform larger-scale targeted functional studies, which may help to better understand the pathogenesis of B. pseudomallei, and subsequently, may improve human health.
A maximum likelihood phylogeny inferred from a concatenation of ~60,000 core-genome SNPs demonstrated that the eleven isolates sequenced in the current study represent broad phylogenetic diversity. The retention index (RI) value, which provides a representation of the homoplasy in the dataset, demonstrated signs of homoplasy, which can confound accurate phylogenetic reconstruction. Plotting the observed homoplasy density (HD) across both chromosomes of B. pseudomallei K96243 demonstrated that the homoplasy was evenly distributed, with no isolated regions of recombination in the core genome. Although this underlying homoplasy may confound phylogenetic relationships, especially in deeply branching nodes, the phylogeny still demonstrates the overall diversity of the eleven isolates sequenced in the current study.
Differences in the distribution of virulence-associated genes were observed based a LS-BSR analysis. One clear difference was the presence of the B. mallei bimA (BimABm) allele in MSHR668 and the B. pseudomallei version (BimABp) in all other isolates (Fig. 3). In previous studies, 12% of Australian isolates contained BimABm [55, 57], although both versions appear to perform actin-based motility effectively. An association between neurological melioidosis and strains with BimABm was recently reported [55]. Severe clinical presentations have been associated with the co-occurrence of BimABm and the hemagglutinin, fhaB3. The lack of fhaB3 in isolates exhibiting BimABm was correlated with cutaneous melioidosis without sepsis [55]. Testing isolates with varied distributions of these virulence components will help corroborate these associations.
The Inv/Mxi-Spa-like type III secretion system (T3SS-3) [58] is essential for the survival of B. pseudomallei in the host [59, 60] and closely resembles secretion systems found in other animal pathogens (Salmonella spp. and Shigella spp.). B. pseudomallei isolates 1026b and HBPUB10134a appear to have reduced homology for BPSS1528, which is described as a (HNS-like regulatory) hypothetical protein in the T3SS-3 system. Several proteins act together to form a pore that becomes bound to the host membrane, thus facilitating the delivery of effector proteins [61, 62]. This system is also likely involved in defenses against autophagy by transporting the BopA effector [63, 64]. In this study, we observed sequence homology variation among many of the isolates in the gene, BPSS1629, from the T3SS-2 cluster.
One of the most dramatic differences observed between isolates was from representative genes in the Yersinia-like fimbriae (YLF) gene cluster and the BTFC gene cluster. This division is mutually exclusive [54, 55] and it is unclear whether one cluster confers enhanced virulence over the other and no correlations have been identified between gene cluster and disease severity [55]. While YLF genes are generally associated with isolates from Thailand [55], we found no geographical correlation in the small sample set that we analyzed in the current study (Fig. 3).
The FDA Animal Rule was set up to identify a set of relevant isolates that could be used in lieu of human clinical trials in the development of effective medical countermeasures against human disease, including melioidosis. The data presented in this study will provide a genomic background to better understand virulence in B. pseudomallei and may also help in the development of more effective medical countermeasures.
Supporting Information
S1 Dataset. The complete LS-BSR matrix for all coding regions in each genome investigated.
https://doi.org/10.1371/journal.pone.0121052.s001
(BZ2)
S2 Dataset. A NASP (http://tgennorth.github.io/NASP/) matrix containing all SNPs from non-duplicated regions from all genomes queried.
https://doi.org/10.1371/journal.pone.0121052.s002
(BZ2)
S1 Fig. Synteny dot plots between finished genomes available in GenBank and draft genomes generated in this study from re-sequencing studies.
Dot plots were generated using the mummerplot method in MUMmer.
https://doi.org/10.1371/journal.pone.0121052.s003
(TIF)
S2 Fig. A Kaplan-Meier curve of survival probabilities based on the BALB/c mice challenge studies conducted in the current study.
The survival probabilities were calculated using the ‘survival’ package in R [12].
https://doi.org/10.1371/journal.pone.0121052.s004
(TIF)
S1 Table. Sequencing information for isolates sequenced in the current study.
https://doi.org/10.1371/journal.pone.0121052.s005
(PDF)
S2 Table. Accession information for reference genomes.
https://doi.org/10.1371/journal.pone.0121052.s006
(PDF)
S3 Table. Virulence associated genes in the current study.
https://doi.org/10.1371/journal.pone.0121052.s007
(PDF)
S4 Table. Survival information over the course of BALB/c challenge studies for all strains challenged.
https://doi.org/10.1371/journal.pone.0121052.s008
(PDF)
Author Contributions
Conceived and designed the experiments: JWS PK AT HCG JMS. Performed the experiments: KEV REC. Analyzed the data: CJA KJC. Contributed reagents/materials/analysis tools: JWS PK KEV BJC. Wrote the paper: JWS.
References
- 1. Wiersinga WJ, Currie BJ, Peacock SJ. Melioidosis. N Engl J Med. 2012;367(11):1035–44. pmid:22970946
- 2. Cheng AC, Currie BJ. Melioidosis: epidemiology, pathophysiology, and management. Clin Microbiol Rev. 2005;18(2):383–416. pmid:15831829
- 3. Currie BJ, Ward L, Cheng AC. The epidemiology and clinical spectrum of melioidosis: 540 cases from the 20 year Darwin prospective study. PLoS neglected tropical diseases. 2010;4(11):e900. pmid:21152057
- 4. Limmathurotsakul D, Peacock SJ. Melioidosis: a clinical overview. British medical bulletin. 2011;99:125–39. pmid:21558159
- 5. Peacock SJ. Melioidosis. Current opinion in infectious diseases. 2006;19(5):421–8. pmid:16940864
- 6. Schweizer HP. Mechanisms of antibiotic resistance in Burkholderia pseudomallei: implications for treatment of melioidosis. Future Microbiol. 2012;7(12):1389–99. pmid:23231488
- 7. Van Zandt KE, Tuanyok A, Keim PS, Warren RL, Gelhaus HC. An objective approach for Burkholderia pseudomallei strain selection as challenge material for medical countermeasures efficacy testing. Front Cell Infect Microbiol. 2012;2:120. pmid:23057010
- 8.
FDA. Guidance for industry animal models—essential elements to address efficacy under the animal rule. Rockville, MD: Department of health and human services, center for drug evaluation and research (CDER) and center for biologics evaluation and research (CBER), 2009.
- 9. Leakey AK, Ulett GC, Hirst RG. BALB/c and C57Bl/6 mice infected with virulent Burkholderia pseudomallei provide contrasting animal models for the acute and chronic forms of human melioidosis. Microb Pathog. 1998;24(5):269–75. pmid:9600859
- 10. Nandi T, Ong C, Singh AP, Boddey J, Atkins T, Sarkar-Tyson M, et al. A genomic survey of positive selection in Burkholderia pseudomallei provides insights into the evolution of accidental virulence. PLoS Pathog. 2010;6(4):e1000845. pmid:20368977
- 11. Sahl JW, Stone JK, Gelhaus HC, Warren RL, Cruttwell CJ, Funnell SG, et al. Genome Sequence of Burkholderia pseudomallei NCTC 13392. Genome Announc. 2013;1(3).
- 12.
R Core Team RCT. R: A language and environment for statistical computing 2013. Available: http://www.R-project.org.
- 13. Kozarewa I, Turner DJ. 96-plex molecular barcoding for the Illumina Genome Analyzer. Methods Mol Biol. 2011;733:279–98. pmid:21431778
- 14. Pop M, Phillippy A, Delcher AL, Salzberg SL. Comparative genome assembly. Brief Bioinform. 2004;5(3):237–48. pmid:15383210
- 15. Assefa S, Keane TM, Otto TD, Newbold C, Berriman M. ABACAS: algorithm-based automatic contiguation of assembled sequences. Bioinformatics. 2009;25(15):1968–9. pmid:19497936
- 16. Tsai IJ, Otto TD, Berriman M. Improving draft assemblies by iterative mapping and assembly of short reads to eliminate gaps. Genome biology. 2010;11(4):R41. pmid:20388197
- 17. Simpson JT, Wong K, Jackman SD, Schein JE, Jones SJ, Birol I. ABySS: a parallel assembler for short read sequence data. Genome Res. 2009;19(6):1117–23. pmid:19251739
- 18. Angiuoli SV, Salzberg SL. Mugsy: Fast multiple alignment of closely related whole genomes. Bioinformatics. 2010.
- 19. Blankenberg D, Taylor J, Nekrutenko A. Making whole genome multiple alignments usable for biologists. Bioinformatics. 2011;27(17):2426–8. pmid:21775304
- 20. Sahl JW, Steinsland H, Redman JC, Angiuoli SV, Nataro JP, Sommerfelt H, et al. A comparative genomic analysis of diverse clonal types of enterotoxigenic Escherichia coli reveals pathovar-specific conservation. Infect Immun. 2011;79(2):950–60. pmid:21078854
- 21. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Mol Biol. 1990;215(3):403–10. pmid:2231712
- 22. Otto TD, Sanders M, Berriman M, Newbold C. Iterative Correction of Reference Nucleotides (iCORN) using second generation sequencing technology. Bioinformatics. 2010;26(14):1704–7. pmid:20562415
- 23. Godoy D, Randle G, Simpson AJ, Aanensen DM, Pitt TL, Kinoshita R, et al. Multilocus sequence typing and evolutionary relationships among the causative agents of melioidosis and glanders, Burkholderia pseudomallei and Burkholderia mallei. J Clin Microbiol. 2003;41(5):2068–79. pmid:12734250
- 24.
Li H. Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM. arXivorg. 2013(arXiv:1303.3997 [q-bio.GN]).
- 25. McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, et al. The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 2010;20(9):1297–303. pmid:20644199
- 26.
Delcher AL, Salzberg SL, Phillippy AM. Using MUMmer to identify similar regions in large sequence sets. Curr Protoc Bioinformatics. 2003;Chapter 10:Unit 10 3.
- 27. Cingolani P, Platts A, Wang le L, Coon M, Nguyen T, Wang L, et al. A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-2; iso-3. Fly. 2012;6(2):80–92. pmid:22728672
- 28. Milne I, Bayer M, Cardle L, Shaw P, Stephen G, Wright F, et al. Tablet—next generation sequence assembly visualization. Bioinformatics. 2010;26(3):401–2. pmid:19965881
- 29. Delcher AL, Phillippy A, Carlton J, Salzberg SL. Fast algorithms for large-scale genome alignment and comparison. Nucleic Acids Res. 2002;30(11):2478–83. pmid:12034836
- 30. Holden MT, Titball RW, Peacock SJ, Cerdeno-Tarraga AM, Atkins T, Crossman LC, et al. Genomic plasticity of the causative agent of melioidosis, Burkholderia pseudomallei. Proceedings of the National Academy of Sciences of the United States of America. 2004;101(39):14240–5. pmid:15377794
- 31. Stamatakis A. RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics. 2006;22(21):2688–90. pmid:16928733
- 32. Stamatakis A. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics. 2014.
- 33. Farris JS. THE RETENTION INDEX AND THE RESCALED CONSISTENCY INDEX. Cladistics. 1989;5(4):417–9.
- 34. Schliep KP. phangorn: phylogenetic analysis in R. Bioinformatics. 2011;27(4):592–3. pmid:21169378
- 35.
Wilgenbusch JC, Swofford D. Inferring evolutionary trees with PAUP*. Curr Protoc Bioinformatics. 2003;Chapter 6:Unit 6 4.
- 36. Krzywinski M, Schein J, Birol I, Connors J, Gascoyne R, Horsman D, et al. Circos: an information aesthetic for comparative genomics. Genome Res. 2009;19(9):1639–45. pmid:19541911
- 37. Galyov EE, Brett PJ, DeShazer D. Molecular insights into Burkholderia pseudomallei and Burkholderia mallei pathogenesis. Annu Rev Microbiol. 2010;64:495–517. pmid:20528691
- 38. Lazar Adler NR, Govan B, Cullinane M, Harper M, Adler B, Boyce JD. The molecular and cellular basis of pathogenesis in melioidosis: how does Burkholderia pseudomallei cause disease? FEMS Microbiol Rev. 2009;33(6):1079–99. pmid:19732156
- 39. 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(4):272–82. pmid:16541135
- 40. Kim HS, Schell MA, Yu Y, Ulrich RL, Sarria SH, Nierman WC, et al. Bacterial genome adaptation to niches: divergence of the potential virulence genes in three Burkholderia species of different survival strategies. BMC Genomics. 2005;6:174. pmid:16336651
- 41. Tuanyok A, Auerbach RK, Brettin TS, Bruce DC, Munk AC, Detter JC, et al. A horizontal gene transfer event defines two distinct groups within Burkholderia pseudomallei that have dissimilar geographic distributions. J Bacteriol. 2007;189(24):9044–9. pmid:17933898
- 42. Tuanyok A, Leadem BR, Auerbach RK, Beckstrom-Sternberg SM, Beckstrom-Sternberg JS, Mayo M, et al. Genomic islands from five strains of Burkholderia pseudomallei. BMC Genomics. 2008;9:566. pmid:19038032
- 43. Sahl JW, Caporaso JG, Rasko DA, Keim P. The large-scale blast score ratio (LS-BSR) pipeline: a method to rapidly compare genetic content between bacterial genomes. PeerJ 2014;2:e332. pmid:24749011
- 44. Rasko DA, Myers GS, Ravel J. Visualization of comparative genomic analyses by BLAST score ratio. BMC Bioinformatics. 2005;6:2. pmid:15634352
- 45. Kruskal WH, Wallis A. Use of ranks in one-criterion variance analysis. Journal of the American Statistical Association. 1952;47(260):583–621.
- 46. Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, et al. QIIME allows analysis of high-throughput community sequencing data. Nature methods. 2010;7(5):335–6. pmid:20383131
- 47. Benjamini Y, Hochberg Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society Series B. 1995;57(1):289–300.
- 48. Benson DA, Karsch-Mizrachi I, Clark K, Lipman DJ, Ostell J, Sayers EW. GenBank. Nucleic acids research. 2012;40(Database issue):D48–53. pmid:22144687
- 49. Ulett GC, Currie BJ, Clair TW, Mayo M, Ketheesan N, Labrooy J, et al. Burkholderia pseudomallei virulence: definition, stability and association with clonality. Microbes Infect. 2001;3(8):621–31. pmid:11445448
- 50. Tumapa S, Holden MT, Vesaratchavest M, Wuthiekanun V, Limmathurotsakul D, Chierakul W, et al. Burkholderia pseudomallei genome plasticity associated with genomic island variation. BMC Genomics. 2008;9:190. pmid:18439288
- 51. Sitthidet C, Stevens JM, Chantratita N, Currie BJ, Peacock SJ, Korbsrisate S, et al. Prevalence and sequence diversity of a factor required for actin-based motility in natural populations of Burkholderia species. J Clin Microbiol. 2008;46(7):2418–22. pmid:18495853
- 52. Kespichayawattana W, Rattanachetkul S, Wanun T, Utaisincharoen P, Sirisinha S. Burkholderia pseudomallei induces cell fusion and actin-associated membrane protrusion: a possible mechanism for cell-to-cell spreading. Infection and Immunity. 2000;68(9):5377–84. pmid:10948167
- 53. Dowling AJ, Wilkinson PA, Holden MTG, Quail MA, Bentley SD, Reger J, et al. Genome-Wide Analysis Reveals Loci Encoding Anti-Macrophage Factors in the Human Pathogen Burkholderia pseudomallei K96243. Plos One. 2010;5(12).
- 54. Tuanyok A, Leadem BR, Auerbach RK, Beckstrom-Sternberg SM, Beckstrom-Sternberg JS, Mayo M, et al. Genomic islands from five strains of Burkholderia pseudomallei. BMC genomics. 2008;9. pmid:18186939
- 55. Sarovich DS, Price EP, Webb JR, Ward LM, Voutsinos MY, Tuanyok A, et al. Variable Virulence Factors in Burkholderia pseudomallei (Melioidosis) Associated with Human Disease. PLoS ONE. 2014;9(3):e91682. pmid:24618705
- 56. Zusman T, Feldman M, Halperin E, Segal G. Characterization of the icmH and icmF genes required for Legionella pneumophila intracellular growth, genes that are present in many bacteria associated with eukaryotic cells. Infection and Immunity. 2004;72(6):3398–409. pmid:15155646
- 57. Sitthidet C, Korbsrisate S, Layton AN, Field TR, Stevens MP, Stevens JM. Identification of motifs of Burkholderia pseudomallei BimA required for intracellular motility, actin binding, and actin polymerization. J Bacteriol. 2011;193(8):1901–10. pmid:21335455
- 58. Holden MTG, Titball RW, Peacock SJ, Cerdeno-Tarraga AM, Atkins T, Crossman LC, et al. Genomic plasticity of the causative agent of melioidosis, Burkholderia pseudomallei. Proc Natl Acad Sci U S A. 2004;101(39):14240–5. pmid:15377794
- 59. Stevens MP, Haque A, Atkins T, Hill J, Wood MW, Easton A, et al. Attenuated virulence and protective efficacy of a Burkholderia pseudomallei bsa type III secretion mutant in murine models of melioidosis. Microbiology-(UK). 2004;150:2669–76. pmid:15289563
- 60. Stevens MP, Wood MW, Taylor LA, Monaghan P, Hawes P, Jones PW, et al. An Inv/Mxi-Spa-like type III protein secretion system in Burkholderia pseudomallei modulates intracellular behaviour of the pathogen. Mol Microbiol. 2002;46(3):649–59. pmid:12410823
- 61. Bleves S, Viarre V, Salacha R, Michel GPF, Filloux A, Voulhoux R. Protein secretion systems in Pseudomonas aeruginosa: A wealth of pathogenic weapons. International Journal of Medical Microbiology. 2010;300(8):534–43. pmid:20947426
- 62. Haraga A, West TE, Brittnacher MJ, Skerrett SJ, Miller SI. Burkholderia thailandensis as a Model System for the Study of the Virulence-Associated Type III Secretion System of Burkholderia pseudomallei. Infection and Immunity. 2008;76(11):5402–11. pmid:18779342
- 63. Gong L, Cullinane M, Treerat P, Ramm G, Prescott M, Adler B, et al. The Burkholderia pseudomallei Type III Secretion System and BopA Are Required for Evasion of LC3-Associated Phagocytosis. Plos One. 2011;6(3).
- 64. Ray K, Marteyn B, Sansonetti PJ, Tang CM. Life on the inside: the intracellular lifestyle of cytosolic bacteria. Nat Rev Microbiol. 2009;7(5):333–40. pmid:19369949