https://orcid.org/0000-0002-5285-0432
Figures
Abstract
Cases of melioidosis have been recorded for many years in Vietnam though it is still not a nationally reportable disease in Vietnam. More research is needed to understand the disease ecology and public health impacts of melioidosis in the country. To this aim, broadening the knowledge base of strains and epidemiology of infections in relation to genotypes present in the soil reservoir can tell us about the propensity of Burkholderia pseudomallei genotypes to transmit from soil to humans. Thirty-five clinical B. pseudomallei isolates, ten from soil, one from swine, and one from a bear were collected by the Institute of Microbiology and Biotechnology, Vietnam National University and sequenced at the National Institute of Hygiene and Epidemiology in Hanoi. The clinical strains were isolated from melioidosis patients from Ha Tinh in each month of 2020 (except July). There were 15 STs identified and four of the clinical isolates were new sequence types (ST) as determined by traditional seven marker multi-locus sequence typing (MLST) analysis. Twenty of the thirty-five (57%) clinical strains isolated in this study were ST 41, with ST 41 isolates obtained throughout the year and across Ha Tinh province with core genome (cg) MLST identifying finer scale differences. ST 41 was recovered from one soil sample approximately 1 year after the clinical isolates. cgMLST analysis and whole genome SNP analysis revealed nucleotide differences among strains in Ha Tinh historically contextualizing them in Vietnam and globally. As melioidosis moves towards a reportable disease in Vietnam, molecular epidemiological methods can connect human, veterinary, and environmental genotypes of concern.
Author summary
The disease melioidosis has been intensely studied in Thailand and Northern Australia while many other locations have faced a lack of investigation, including other countries in Southeast Asia. Among the lesser studied areas is Vietnam. The bacterium that causes melioidosis, Burkholderia pseudomallei, can cause severe infections and is a public health concern in wet tropical locations around the world. This study takes a detailed molecular epidemiological look at clinical, soil, and veterinary B. pseudomallei isolates from North Central Vietnam. The isolates were whole genome sequenced which revealed STs shared among soil, veterinary, and clinical B. pseudomallei strains. Clinical sequence types were heavily enriched for isolates that shared unique gene sets that could indicate geographically specific bacterial virulence mechanisms.
Citation: Norris MH, Au La TH, Metrailer MC, Viet Nguyen H, Thi Le Tran Q, Jiranantasak T, et al. (2026) Expanding the molecular epidemiology of melioidosis in North Central Vietnam. PLoS Negl Trop Dis 20(2): e0013945. https://doi.org/10.1371/journal.pntd.0013945
Editor: Stuart D. Blacksell, Mahidol Univ, Fac Trop Med, THAILAND
Received: September 9, 2025; Accepted: January 19, 2026; Published: February 9, 2026
Copyright: © 2026 Norris 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 the data from this work is available in the manuscript itself or in publicly available databases at SRA BioProject PRJNA1180080.
Funding: Funding for this work was provided by US Defense Threat Reduction Agency grant HDTRA12010003 to M.H.N, H.T.T.N, T.T.T., and J.K.B. The funders had no role in the 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
The disease melioidosis, caused by the bacterium Burkholderia pseudomallei, was first described in 1912 by A. Whitmore and C.S. Krishnaswami as a disease afflicting injectional morphine users in what was then Rangoon, Burma (now Yangon, Myanmar) [1]. It occurs across the globe in wet tropical environments with an evolving epidemiology [2–4]. The disease has been intensely studied in Thailand and Northern Australia [5–8] while many other locations have faced a lack of investigation, including other countries in Southeast Asia. Among the lesser studied areas is Vietnam, a country where melioidosis cases were associated with American soldiers returning from war thus earning the moniker “Vietnamese time bomb”. Melioidosis cases from travel to Vietnam have also been recorded [9]. Even with the preponderance of evidence that melioidosis exists in Vietnam, the true incidence of disease in humans, the spatial context of disease, and the genetics of strains present in the country remain to be fully understood.
B. pseudomallei genetic diversity has been studied with seven-locus multi-locus sequence typing (MLST) [10]. MLST analysis allows differentiation of the 2,220 sequence types (STs) currently identified. More recently MLST schemes targeting genes in the core genome of B. pseudomallei have been developed [11] where whole genome sequences are used for allelic typing at over four thousand genome loci. Currently there are more than 1,000 core genome sequence types (cgSTs) deposited in the public pubMLST database [12]. While vital epidemiological information can be derived from comparing cgSTs, availability of whole genome sequence allows single nucleotide polymorphism (SNP) analysis for highly detailed phylogenetic analysis. Burkholderia pseudomallei genetic variation has been intensely studied with data heavily skewed to Thailand and Northern Australia. The total number of B. pseudomallei isolates deposited in the seven-marker MLST sequence typing PubMLST database total 7,741. Of those isolates 1,163 and 3,903, from Thailand and Australia respectively, as of December 2025. The remaining ~34% of isolates are from the rest of the world, with 269 from Vietnam (inclusive of this study). PubMLST also hosts whole genomes. Of the 1,272 B. pseudomallei genomes deposited as PubMLST genomes, 874 are from Australia, 41 are from Thailand, and 55 are from Vietnam. In the NCBI sequence read archive there are 9,799 entries for B. pseudomallei. Australia has 1,760, Thailand 3,402 and Vietnam 139 (inclusive of this study). These numbers illustrate the overwhelming proportion of B. pseudomallei strain typing and sequencing done in Australia and Thailand, leaving much to be discovered in other countries where melioidosis is known to be endemic yet under studied, as in Vietnam [13]. A few molecular epidemiological studies of strains in Vietnam have been completed. Seven marker MLST analysis of four strains of B. pseudomallei collected in 1964 was an early, limited study on strain diversity in Vietnam [14]. More recent studies in Vietnam utilized genomic epidemiology to link a single human case to soil isolates in Nghe An province [11] and identified a contaminated bore-well as the source of B. pseudomallei that killed three children in Hanoi [15].
Geographic biases can affect epidemiology of melioidosis by limiting place-specific bacterial diversity and obscuring impact of patient ethnicity, while simultaneously ignoring differences in ecology and culture that play important roles in disease. The current work is the outcome of a multi-institutional project to increase the diagnosis of B. pseudomallei in Vietnam. By enhancing awareness and microbiological methods of diagnosis, new strains could be identified, isolated and whole genome sequenced. The more strains with whole genome sequence available the more details regarding strain diversity, transmission to humans and animals, and environmental persistence can be gained. To that end, increased capacity for diagnostic isolation of B. pseudomallei strains was implemented to improve the molecular epidemiology of melioidosis in the North Central Vietnamese province of Ha Tinh. The enhanced diagnostic techniques were employed in Ha Tinh Provincial General Hospital, and the increase in positively identified melioidosis samples throughout 2020 provided a series of clinical isolates for genome sequencing and molecular epidemiological analysis. Follow up environmental sampling and isolation of B. pseudomallei from soils in Ha Tinh provided genetic snapshots of bacterial diversity maintained in the soil. To complete our One-Health approach to characterizing melioidosis in Vietnam, two veterinary isolates, one from a pig and one from a bear, were sequenced. As the reporting of incidence increases in Vietnam, it is anticipated the disease will become nationally reportable allowing development and implementation of country wide public health policies around prevention, diagnostics, and treatment of melioidosis.
Methods
Ethics statement
Clinical cases were diagnosed during routine hospital procedures. The clinical data were retrieved based on bacterial cultures for this retrospective analysis. Medical records were deidentified and correlated with diagnostic microbiology findings. The retrospective studies of human medical records and microbiological isolates were approved by the University of Florida Institutional Review Board under protocol IRB# IRB202301112 and by the Institutional Review Board at the University of Medicine and Pharmacy, Vietnam National University under protocol IRB# IRB00013221. Animal sampling took place following Article 21 under the Vietnam National Law governing animal welfare. Swine sampling was performed by the Vietnam Institute of Animal and Veterinary Science under the guidance of the Ministry of Agriculture and Rural Development Circular 7.
Study location, clinical records and diagnosis
Ha Tinh Province is located in North Central Vietnam and Ha Tinh Provincial General Hospital is the provincial hospital that accepts patients primarily from the districts within Ha Tinh province (Fig 1A). A major component of the current project was increasing local knowledge and diagnostic capacity for melioidosis in hospitals of North Central Vietnam. This was accomplished by introducing more sensitive diagnostic techniques while educating laboratorians and clinicians through workshops and training. Increased isolation rates of B. pseudomallei from patients in Ha Tinh Provincial General Hospital allowed for molecular analysis of a series of clinical isolates across patient symptom presentations and syndromic outcomes at various locations in Ha Tinh Province (Fig 1B). The research collaboration allowed for rapid screening and molecular characterization of B. pseudomallei clinical isolates identified during routine patient care. Patients with high fevers and suspected melioidosis had their diagnostic blood samples processed for B. pseudomallei as previously described [16]. Briefly, 1 ml of patient whole blood was injected into blood bottles (BACT/ALERT FA Plus, Biomerieux, Marcy-I’Étoile, France) and incubated at 37°C until growth was observed. Bottles with growth were struck on to Ashdown’s selective agar and incubated at 37°C for up to 48 h [17]. Suspect colonies were Gram-stained and genomic DNA was extracted from colonies with rod-shaped Gram-negative bacteria. Positive identification of B. pseudomallei was confirmed using the TTS1-orf2 quantitative polymerase chain reaction (qPCR) assay [18]. All province and district names and their administrative levels mentioned in this study refer to those in use before July 1st, 2025. This work was performed under University of Florida IRB# IRB202301112 and the University of Medicine and Pharmacy, Vietnam National University IRB# IRB00013221.
A) Provinces are colored differently while the smaller subdivisions are districts. The provinces Vinh Phuc, Nghe An, and Ha Tinh are labeled. The white circle with the black dot indicates the location of Hanoi, Vietnam where IMBT and NIHE are located and the red plus sign indicates the location of Ha Tinh Provincial General Hospital in Ha Tinh province where the clinical and environmental strains were isolated. B) A map showing locations of clinical B. pseudomallei strains from Ha Tinh province. Clinical strains (white points) isolated in 2020 and environmental B. pseudomallei strains (gray points) from 2016 and 2022 in relation to Ha Tinh Provincial General Hospital (red cross) within Ha Tinh Province. White lines indicate districts within Ha Tinh. The names of provinces and district-level administrative divisions are those used before July 1st, 2025. Regional country basemaps (https://gadm.org/maps.html), country (https://purl.stanford.edu/dp747xv3917), province (https://purl.stanford.edu/mq929cm3505) and district (https://purl.stanford.edu/zq121fy4008) level base maps were used following the GADM license (https://gadm.org/license.html).
Veterinary diagnosis
Two bacterial strains from animals were also sequenced during this analysis, including one from a pig in Nghe An, located near Ha Tinh in central Vietnam, and one from an Asiatic brown bear in Vinh Phuc, a mountainous province to the northwest of Hanoi [19]. Tracheal swabs from an underweight swine serologically positive by ELISA were plated on Ashdown’s agar to isolate colonies [20,21]. For the bear isolate, a pus swab from a lethargic bear was cultured at Central Hospital 74 in Vinh Phuc on routine media, including blood agar and MacConkey agar. After 2-days incubation at 37°C, growth was observed. In each case, isolated colonies with the B. pseudomallei phenotype were plucked and saved for qPCR confirmation and whole genome sequencing. Animal sampling took place following Article 21 under the Vietnam National Law governing animal welfare. Swine sampling was performed by the Vietnam Institute of Animal and Veterinary Science under the guidance of the Ministry of Agriculture and Rural Development Circular 7. The bear sample was collected by rural veterinarians following article 21 then transported to IMBT.
Environmental sampling and bacterial isolation
Regional environmental samples from 2016 and 2022 (before and after the clinical cases) were from northern and central areas of Ha Tinh province where the cases were reported. In 2016, 80 samples at 16 randomly chosen sites were collected. In 2022, 50 soil samples at 10 sites in the Central Ha Tinh region proximate to areas with high case reports from 2020 and 50 soil samples at 10 sites where less melioidosis was reported in 2020 (Northeastern Ha Tinh) were collected. Soil samples were obtained from 30 cm below the surface using a JMC Soil Samplers soil sampling auger. At each sampling site, five soil samples were taken, with 5–10 m between each other. The steel auger used for collection was cleaned with bottled water and disinfected with 70% (v/v) ethanol after each sampling point. Approximately 200 g of soil sample was collected in plastic ziploc bags and transferred to the Institute of Microbiology and Biotechnology (IMBT), Vietnam National University, Hanoi. Samples were transported and stored at ambient temperature before processing within 1 week for B. pseudomallei isolation using the two-step enrichment method as previously described [22]. Briefly, 10 g of soil was transferred into 50-ml tube containing 20 ml TBSS-C50 broth. After vigorously vortexing, the tube was incubated statically at 40°C for 2 days. Then, 1 ml of the first culture supernatant was transferred to a new 50 ml tube containing 9 ml of erythritol medium (EM). The tube was incubated at 40°C for 4 days prior to plating on Ashdown’s agar. After incubation at 40°C for 4 days, suspected B. pseudomallei colonies were picked up and stored at -70°C in Luria-Bertani broth supplemented with 20% glycerol for further identification and genomic analysis.
Mapping
GPS locations of clinical cases were recorded at diagnosis. The locations are confidential, and these data are not releasable under IRB regulations. For this study, patient locations were protected by dithering all points by 100 meters using the point displacement function in QGIS [23]. GPS coordinates for soil samples were also recorded, dithered, and mapped using QGIS. The names of provinces and district-level administrative divisions (provinces, districts) were those in place prior to July 1, 2025. The shapefiles used in this work are from the Global Administrative Areas 2015 (v2.8) dataset used under the CC BY 4.0 license and were accessed via the following urls on December 31, 2024: administrative level 0 (country) https://purl.stanford.edu/dp747xv3917, administrative level 1 (provinces) https://purl.stanford.edu/mq929cm3505, and administrative level 2 (districts) https://purl.stanford.edu/zq121fy4008. These data are from the GADM dataset which is freely available for publication (https://gadm.org/license.html). Satellite imagery was accessed on December 8, 2025 from the QGIS quick map services EOX::Maps Sentinel-2 cloudless map, which is sourced from the European Space Agency’s Sentinel-2 and publishable under CC BY 4.0.
Bacterial isolation and qPCR confirmation
Burkholderia pseudomallei strains were isolated from patients visiting Ha Tinh Provincial General Hospital during 2020, from the two animals, and soils in 2016 and 2022. The genomic DNA was purified using the Promega Wizard Genomic DNA extraction kit and qPCR confirmed using the TTS1 qPCR assay [18,24] at IMBT at Vietnam National University.
DNA sequencing
Libraries were prepared as previously described [25,26]. Briefly, sequencing libraries were generated using the Illumina Nextera XT DNA library preparation kit according to the manufacturer’s recommendations. DNA sequencing was performed at the National Institute of Hygiene and Epidemiology (NIHE) in Hanoi on the Illumina MiSeq system. The fastq files of the 47 strains sequenced have been deposited in the NCBI Sequence Read Archive under SRA BioProject PRJNA1180080.
Bioinformatics and bacterial typing
Raw data were processed on the University of Hawai’i KOA High Performance Computing Cluster. Whole genome sequencing data were used to generate assembled contigs using the nextflow(nf)-core pipeline bacass [27,28]. This pipeline trims the raw reads based on quality using FastP [29] followed by read assembly with Unicycler [30] and annotation with Prokka [31]. Available fastq files from the NCBI sequence read archive (SRA) for 1468 strains were downloaded, assembled, and annotated by the same pipeline. Strains and metadata are indicated in S1 Table. The assembled genomes were typed by in silico standard MLST and cgMLST analysis using loci allele sequences and schema from pubMLST. Briefly, the B. pseudomallei K96243 genome was used to create a prodigal training file with chewBBACA [32]. All allele sequences for each locus were downloaded from pubMLST by API and allele calling was performed using the Coreugate nextflow pipeline [33]. The cgMLST profiles from the strains sequenced in this work were submitted to pubMLST and had cgSTs assigned for the comparison to strains in the database. Whole genome SNP analysis was used to identify molecular genetic patterns and to identify diagnostic targets. It was performed by analyzing the assembled whole genomes with PhaME, a pipeline developed at Los Alamos National Lab (LANL) [34]. Briefly, draft assemblies were aligned to the B. pseudomallei 1026b reference (GCA_000260515) using nucmer. SNP and gap locations are used to generate core genome alignments. Phylogenetic tree inference and ultrafast bootstrapping were performed with IQ-TREE by implementing the transversion with estimated base frequencies, invariant sites, gamma rate variation, model and incorporation of the proportion of invariable sites model (IQ-TREE option -m TEST found TVM + F + I + G4 was the best model as determined by AIC and BIC testing) [35]. The phylogenetic trees in Newick format were visualized in the standalone grapetree [36] module and with iTOL [37].
Accessory gene analysis
Gene annotations strains were analyzed by the Roary pan-genome analysis pipeline [38] to identify genes or operons unique to the twenty-one ST 41 B. pseudomallei clinical isolates and one environmental isolate identified in this study. GFF3 files generated by Prokka from assemblies in this study were compared to forty-one closely related non-ST 41 strains downloaded from GenBank using default Roary parameters. The nucleotide sequences encoding for the proteins present only in ST 41 strains were extracted. Blastn of these sequences using the default settings with a 0.05 expected mismatch threshold was used to confirm absence or presence of these gene regions in other bacterial strains, including other B. pseudomallei. The gene region was considered as present if alignment coverage was > 90% and there was > 90% nucleotide identity.
Results
Burkholderia pseudomallei was isolated from 35 melioidosis patients; 11 that had diabetes (31.4%) (Table 1). Eight of the patients died, eleven patients recovered, and the remaining were lost to follow up after transfer from the hospital. Initial diagnoses included septicemia, soft tissue abscesses, arthritis, pneumonia, pleural effusion, and a urinary tract infection. Burkholderia pseudomallei was isolated in blood from 26 of the patients, pus from seven, sputum from one, and pleural fluid from the last. Whole genome sequencing at NIHE in Hanoi allowed in silico sequence typing. Traditional seven-marker MLST identified nine STs total, including two new STs, among the clinical isolates. Surprisingly, 20 out of 35 strains (57%) isolated from the 35 patients across the year were ST 41. All other clinical STs were unique. Four of the clinical isolates were new STs. Strain 34HT was assigned ST 2215, 43HT was assigned ST 2216, and 10HT was assigned ST 2220. The fourth strain, 46HT matched ST 41 except the lipA sequence was not found. This region of the genome was not present, indicating a deletion encompassing the lipA locus. Whole genome SNP analysis and cgMLST placed this strain in the ST 41 cluster even though the lipA sequence was missing, bringing the total ST 41 strains to 21 out of 35 (60%). The earliest ST 41 strain, isolate 2002721102 from a 1983 human melioidosis case, was deposited in pubMLST by the US CDC, likely a travel related infection from Southeast Asia possibly even Vietnam [39]. During the preparation of this manuscript, the US CDC published a study revealing environmentally acquired US melioidosis cases occurring in 2024 in the state of Georgia [40]. The 2024 cases were linked spatially and genetically to the above mentioned 1983 melioidosis case. The closest related strains from this current work are 45HT from Cam Xuyen district and 48HT from Thach Ha district. Genetically they fall outside the main clades where the Georgia strains are located but indicate a potential common ancestor from Central Vietnam.
Soil samples in Ha Tinh from 2016 and 2022, four years before and again two years after the clinical series, were processed for B. pseudomallei isolation to provide molecular environmental context to bacterial diversity in the area. Burkholderia pseudomallei strains were successfully isolated from 15 soil samples at eight randomly chosen sites from 2016 (S1 Fig and S2 Table). In 2022, 50 samples from 10 sites in an area coinciding with 2020 high human incidence were isolated in Central Ha Tinh. Burkholderia pseudomallei was successfully isolated from eight of ten of those sites (19 of 50 samples; S2 Table). Another 50 soil samples were isolated in Northeastern Ha Tinh where few human cases were reported. One sample of 50, at only 1 of 10 sites, were positive from this sampling area. In total, B. pseudomallei was isolated from 20 soil samples at 9 sites. A subset of 10 of 35 isolates were sequenced in this work, three from 2016 and seven from 2022. Two ST 56 strains were isolated in 2022. This sequence type has been found in the environment and in human infections across Southeast Asia, including Bangladesh, Thailand, Cambodia, Malaysia, and Vietnam. Among the B. pseudomallei soil isolates from the 2016 and 2022 sampling Ha Tinh events, two STs overlapped with the clinical STs. These two STs found in human patients and also the soil sampled during 2022 were STs 41 (strain SHT3.5) and 307 (strain SHT4.5G) (Table 2). Two of the soil isolates, both from 2016, had novel STs. Strain MT22.4 was assigned ST 2217 and strain MT27.4.1 was assigned ST 2218 by pubMLST. Burkholderia pseudomallei isolate VP74 G20.1 from the bear in Vinh Phuc had a novel ST assigned, ST 2219, while swine isolate TY01 from Nghe An was ST 55. Sequence type 55 was found in both the 2016 swine from Nghe An (strain TY01) and a 2022 Ha Tinh soil sample (strain SHT8.1). Other ST 55 strains in the pubMLST database have been isolated from clinical melioidosis in China and Malaysia. One ST 541 strain, MT 24.1, was isolated from soil. This ST was associated with a cluster of human deaths in Northern Vietnam and traced back to a contaminated borewell [15].
Whole genome SNP analysis of 1,468 B. pseudomallei strains from Australia and Southeast Asia, including the 47 from this work, by PhAME showed phylogenetic linkages of the isolates. For a comparison of typing methodologies, cgMLST analysis was performed on these strains using a previously published cgMLST scheme [11] and included isolates from Australia, Thailand, Laos, and Vietnam. SNP (Fig 2A) and cgMLST (Fig 2B) analysis revealed similar genetic relationships between the two methods. A majority of the B. pseudomallei sequenced in this work clustered towards the top end of the tree near Australian strains; considered the ancestral, more diverse B. pseudomallei. Local branching was conserved whether the tree was constructed by SNP or cgMLST analysis. Placement of more ancestral branches was different between SNP and cgMLST methods as indicated by shifts in tree location of strains in this work, with pink lines connecting strain leaves in trees made using the two different methods. Statistical bootstrapping was more supportive of strains placed closer together on the tree. The placement of some larger clades was not well supported, reflective of some of the differences in strain placement between SNP and cgMLST trees. The twenty-one ST 41 clinical isolates, were clustered closely to the ST 41 soil isolate SHT 3.5 (Fig 2A, red-dashed box). Both SNP and cgMLST identified two other clusters of clinical melioidosis strains and environmental isolates. One cluster includes soil isolate MT 22.4 and clinical isolates 10HT and 27HT. Another includes clinical isolate 57HT and the two soil isolates SHT 4.5G and SHT 9.1. The B. pseudomallei strain isolated from the bear, VP74 G20.1, was located near a cluster of B. pseudomallei from Vietnam as determined by wgSNP analysis. Other B. pseudomallei strains from our study that were phylogenetically related to bear isolate VP74 G20.1 were strains 16HT and 60HT isolated from human melioidosis. Core genome MLST placed this strain near several from Laos and somewhat further from the aforementioned human clinical isolates. The swine isolate TY01 from 2016 shared similar locations on both trees and, unsurprisingly, fell very close to soil isolate SHT 8.1 from 2022 since both were ST 55.
Whole genome assemblies from 1468 B. pseudomallei strains including the 35 clinical, 10 environmental, and 2 animal isolates sequenced in this study (red branches) were analyzed by whole genome SNP analysis (A, left). High value (>90%) bootstrapped branches as determined using the ultrafast bootstrapping setting in IQ-TREE are colored blue. The PhAME analysis was compared to cgMLST analysis (B, right). The color strip indicates country of isolation. Australia, blue; Laos, green; Thailand, yellow; Vietnam, red. Between the two trees are columns indicating strain source including animal, clinical, or environmental. In the very center ST 41 strains are indicated by red dots and labeled in the dashed box. The locations of strains are labeled and indicated by blue arrows. The pink lines connecting A) and B) compare the location of the sequenced strains between the two trees. Tree scale in A) is in substitutions per site and in B) is cgMLST allele loci differences.
Twenty-one of the thirty-five clinical isolates (60%) identified in this work were ST 41. These ST 41 strains were isolated from patients residing at different locations in Ha Tinh province during the whole study year of 2020 (Fig 3A, red circles). The red circles on the map indicate their varied spatial locations in comparison to the location of Ha Tinh Provincial General Hospital. A closer look at ST 41 on the SNP tree revealed the highly clonal nature of the strains in the ST 41 cluster compared to other related clades (Fig 3B). In some cases, only a few SNPs in the core genomes separated the strains (Fig 3C). Strain SHT 3.5 was the ST 41 isolated from soil 2 years after the 2020 series of cases were diagnosed and was most closely related to strain 05HT from a melioidosis case in Cam Xuyen district, Ha Tinh province. The closely related strains, 40HT, 47HT, and 32HT, located on the distal end of the ST 41 SNP branch were also isolated from melioidosis cases in Cam Xuyen district. A minimum spanning tree (MST) produced using the cgMLST profiles (Fig 4A) of the ST 41 strains showed a phylogeographic correlation among genetically related strains and their geographic location of isolation from the southeast to the northwest of the study area (Fig 4B). There was not a clear temporal correlation between date of infection and strain location on the phylogenetic tree.
A) A map of Ha Tinh province indicating the location of the clinical isolates from 2020 (white circles) and environmental isolates from 2016 and 2022 (black and gray points, respectively) recovered in relation to Ha Tinh Provincial General Hospital (red cross). The points indicated by red circles show which isolates were ST 41 strains. B) Whole genome SNP phylogenetics allowed further resolution of ST 41 strains recovered throughout the year across Ha Tinh province. Red branches indicate strains isolated and whole genome sequenced in this study. The color strip indicates country of origin. C) A zoomed in section of the SNP tree showing SNP differences in the core genomes of the ST 41 strains. Gray squares indicate during this study. One ST 41 soil isolate, SHT3.5, was recovered 2 years after the 2020 infection series. Province (https://purl.stanford.edu/mq929cm3505) and district (https://purl.stanford.edu/zq121fy4008) level base maps were used following the GADM license (https://gadm.org/license.html).
A) A minimum spanning tree was created from the B. pseudomallei ST 41 cgMLST profiles. Strain names are located in the circles. The green filled circle is the only ST 41 soil isolate SHT 3.5 from 2022. The remainder are the twenty-one clinical ST 41 isolates. The unfilled circle is strain 46HT that lacked a lipA sequence but still fell in the ST 41 clade. The strains on the tree are connected to their location of isolation in Ha Tinh province by dashed lines. B) The map of Ha Tinh province (highlighted in green) showing the ST 41 strain isolation points circled in red. Country (https://purl.stanford.edu/dp747xv3917), province (https://purl.stanford.edu/mq929cm3505) and district (https://purl.stanford.edu/zq121fy4008) level base maps were used following the GADM license (https://gadm.org/license.html).
The finding that ST 41 B. pseudomallei were isolated from a majority of clinical cases in Ha Tinh province during the study period prompted a closer look at the genetic signatures that were unique to these strains. In case there were some virulence associated factors involved in what could be an ongoing epidemic of ST 41 melioidosis in the study region, a gene-by-gene BLAST was applied to all ST 41 strains in this work, other available ST 41 strains, and a selection of non-ST 41 B. pseudomallei to identify DNA regions found in ST 41 strains and a limited number of other Burkholderiae (Table 3). The largest region of ST 41-specific DNA (region 1) is ~ 15.4. kbp and encodes a PhoH-like and an RpoD sigma factor (Fig 5). Additional genes in this module encode a XerC tyrosine recombinase and two phage integrase proteins, suggesting this region may be the result of a lysogenic phage insertion. Both PhoH and RpoD-like proteins indicate the additional region could impact gene regulation and virulence. Nucleotide level homology results found a region of 99.95% DNA sequence identity across 100% of the module sequence in B. pseudomallei strain HBPUB10134a. This strain was isolated from a human infection in Thailand and has been associated with higher virulence in animal models [41,42]. Burkholderia aenigmatica strain CMCC(B)23010 had 86.40% nucleotide identity across 99% of the region while Paraburkholderia hospita strain mHSR1 had 81.63% nucleotide identity across 81.63% of the region. Burkholderia aenigmatica is a newly named species part of the Burkholderia cepacia complex [43] and P. hospita is a root associated bacteria that can induce resistance against bacterial plant pathogens [44,45]. Other alignments had coverage lower than 50%. Burkholderia pseudomallei strain HBPUB10134a is ST 228 and has some phylogenetic relationship with the ST 41 strains. Three other single gene regions present in ST 41 strains and no other B. pseudomallei encode hypothetical proteins. These regions were present in all ST 41 identified, even the Ha Tinh B. pseudomallei soil isolate SHT 3.5 from 2022, and could be developed as a molecular target to quickly differentiate ST41 and other strains that have this additional module.
The 15,439 bp region of DNA found in ST 41 and at least one other B. pseudomallei strain (high virulence strain HBPUB10134a). Region 1 contains genes encoding in order, XerC tyrosine recombinase, two phage integrase proteins, helix-turn-helix DNA binding domain protein, a cytosine specific DNA methylase, helicase, hypothetical protein, a PhoH-like protein, and RNA polymerase sigma factor RpoD. This region is present in all ST 41 strains found in this work and others.
Discussion
Melioidosis is still not a reportable disease in Vietnam. Education and training have increased the knowledge and diagnosis of the disease in highly endemic regions. This study focused on sequencing and typing of strains at Ha Tinh Provincial General Hospital in Ha Tinh province located in North Central Vietnam following a concerted effort to increase awareness of human melioidosis. The blood of septicemic and pneumonic patients was the most frequent source material for isolating B. pseudomallei. Soil sampling efforts in the region from before and after the series of human cases provided a historical environmental isolate perspective, allowing contextual genetic snapshots of the human cases. Two animal isolates, one from a bear in Northern Vietnam and one from a pig in North Central Vietnam, provided a glimpse at strain types found in atypical and typical animal hosts.
Establishment of a sequencing pipeline in Vietnam was used to connect hospitals in endemic regions to academic and public health centers in Hanoi. Whole genome sequencing provided substantially more data than qPCR and allowed for detailed molecular epidemiological analysis traditional MLST cannot provide. SNP and sequence typing analysis, whether traditional seven marker MLST or several thousand marker cgMLST, can be extracted from the whole genome data and selectively applied depending on the goals. There was no direct correlation between STs in humans and those found in the environment. This is reflective of acquisition of bacteria from the environment, selection in the host, ST- specific pathogenic attributes and the genetic diversity of the pathogen in the soil niche. In this work, traditional MLST found that 60% of the clinical strains isolated from melioidosis patients at Ha Tinh Provincial General Hospital were a single ST, ST 41. The tendency for B. pseudomallei to have a large and expanding accessory genome by incorporating extraneous DNA by natural transformation and recombination can affect core genome-based phylogenetics (like wgSNP and cgMLST). The relatively large proportion of accessory genes limits the size of the core genome and inhibits accurate model determination during phylogenetic reconstruction [46]. This leads to uncertainty in phylogenetic branch placement like that observed between wgSNP and cgMLST trees. However, the high resolution of these techniques is still valuable where statistically supported. SNP and cgMLST analysis agreed with traditional MLST and the higher resolution analysis confirmed the clonal nature of the ST 41 strains. Some strains differed by only a few SNPs or cgMLST loci. The cgMLST minimum spanning tree comparison to location of the isolated strains showed a spatial correlation to the high-resolution genetic typing of the ST 41 strains. Even though enrichment of ST 41 strains was occurring in the patients, their genetics across Ha Tinh province could be generalized from the southwest to the northeast. The lack of infection foci was not surprising given the endemicity of B. pseudomallei in the soils of North Central Vietnam. There could be limited point-source infections within the case data but determining linkages without more epidemiological information is not possible.
Closer analysis of ST 41 strains found genetic regions that could explain the enrichment of this ST in the clinical series during 2020. Only one of the 10 soil isolates in this study was ST 41. Region 1 genes (Table 3) included phage integrase genes, supporting the hypothesis that a lysogenic phage transfer played a role in transfer of this gene module to ST 41 B. pseudomallei from a currently uncharacterized B. pseudomallei strain or possibly from near-neighbor like Burkholderia cepacia. Natural transformation is another mechanism by which naked DNA can be integrated into the genome of highly recombination proficient B. pseudomallei [47]. The predicted function of other genes in this genetic region supports a role for Region 1-encoded genes in human infections. PhoH-like proteins are responsible for upregulating virulence factors in response to phosphate starvation [48–50]. The human host is considered low in phosphate [49] and it is plausible that this additional PhoH protein could affect virulence mechanism induction. Adjacent to the phoH gene is an rpoD homologue. RpoD is a sigma factor (SigA or s70) that is known to activate housekeeping gene regulons important for exponential growth [51,52]. Another DNA binding protein in this region is a helix-turn-helix domain protein. With an abundance of gene regulatory elements present in this ~15 kbp region, Region 1 could be contributing to ST 41 infection rates in Ha Tinh and other regions including the US state of Georgia and is the subject of future investigations.
Conclusions
This one-year provincial level cross-sectional study of human clinical cases at Ha Tinh Provincial General Hospital increases the molecular epidemiological data of melioidosis in an endemic region of Vietnam that has not been well studied. The 47 human, soil, and veterinary B. pseudomallei isolates sequenced as part of this work expands the whole genome data from strains in Vietnam. While the sample numbers in this study are small, interesting molecular epidemiological details were still obtained. A major B. pseudomallei human melioidosis causing sequence type was found during soil sampling. The animal samples currently have limited epidemiological value because only one from each species was identified, however, these data can provide a basis should future outbreaks among domesticated animals or wildlife occur. Limitations of this work include that exploratory soil sampling in southern, central, and northern Ha Tinh predated the clinical survey by several years, 2016 versus 2020, while follow-up sampling occurred in 2022. Despite asynchronous sampling, ST 55 was found in the swine from 2016 and in soil from 2022. Clinical ST41 from 2020 was found in soil from 2022. These examples reflect the long-term environmental stability of B. pseudomallei. Contemporaneous soil and clinical sampling could have generated a more correlated dataset. The current work provides a template for understanding the molecular epidemiology of melioidosis as diagnostic and training programs are moved further south of Ha Tinh into the South Central and Southern regions of Vietnam. Detailed molecular data of the B. pseudomallei in Vietnam is required as the foundation for place-specific diagnostic and vaccine development. Future work encompassing more dedicated soil, veterinary, and clinical studies are required to fully understand the role B. pseudomallei genetics are playing in human infections in Vietnam. This study represents an important step in moving B. pseudomallei towards a reportable disease in Vietnam. Partnering microbiology, clinical, veterinary, and public health representatives will be vital to understanding all facets of melioidosis in Vietnam.
Supporting information
S1 Fig. Map of all soil sampling sites in Ha Tinh.
https://doi.org/10.1371/journal.pntd.0013945.s001
(TIFF)
S1 Table. SRA and strain metadata for strains used to build the phylogenetic trees in Fig 2.
https://doi.org/10.1371/journal.pntd.0013945.s002
(XLSX)
S2 Table. Details of soil sampling sites coordinates, positivity rates, and year.
https://doi.org/10.1371/journal.pntd.0013945.s003
(XLSX)
Acknowledgments
We would like to thank Mr. Nguyen Nhu Hung at Central Hospital 74 in Vinh Phuc province for providing the B. pseudomallei strain isolated from the bear.
References
- 1. Whitmore A, Krishnaswami CS. An account of the discovery of a hitherto undescribed infective disease occuring among the population of Rangoon. Indian Medical Gazette. 1912;1912:262–7.
- 2. Cheng AC, Currie BJ. Melioidosis: Epidemiology, Pathophysiology, and Management. Clin Microbiol Rev. 2005;18:383–416.
- 3. Currie BJ, Meumann EM, Kaestli M. The Expanding Global Footprint of Burkholderia pseudomallei and Melioidosis. Am J Trop Med Hyg. 2023;108(6):1081–3. pmid:37160279
- 4. Limmathurotsakul D, Golding N, Dance DAB, Messina JP, Pigott DM, Moyes CL, et al. Predicted global distribution of Burkholderia pseudomallei and burden of melioidosis. Nat Microbiol. 2016;1(1).
- 5. Seng R, Chomkatekaew C, Tandhavanant S, Saiprom N, Phunpang R, Thaipadungpanit J, et al. Genetic diversity, determinants, and dissemination of Burkholderia pseudomallei lineages implicated in melioidosis in Northeast Thailand. Nat Commun. 2024;15(1):5699. pmid:38972886
- 6. Kaestli M, Grist EPM, Ward L, Hill A, Mayo M, Currie BJ. The association of melioidosis with climatic factors in Darwin, Australia: A 23-year time-series analysis. J Infect. 2016;72(6):687–97. pmid:26945846
- 7. Currie BJ, Mayo M, Ward LM, Kaestli M, Meumann EM, Webb JR, et al. The Darwin Prospective Melioidosis Study: a 30-year prospective, observational investigation. Lancet Infect Dis. 2021;21(12):1737–46. pmid:34303419
- 8. Rachlin A, Mayo M, Webb JR, Kleinecke M, Rigas V, Harrington G, et al. Whole-genome sequencing of Burkholderia pseudomallei from an urban melioidosis hot spot reveals a fine-scale population structure and localised spatial clustering in the environment. Sci Rep. 2020;10: 5443.
- 9. Gauthier J, Gérôme P, Defez M, Neulat-Ripoll F, Foucher B, Vitry T, et al. Melioidosis in Travelers Returning from Vietnam to France. Emerg Infect Dis. 2016;22(9):1671–3. pmid:27532771
- 10. 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
- 11. Lichtenegger S, Trinh TT, Assig K, Prior K, Harmsen D, Pesl J, et al. Development and Validation of a Burkholderia pseudomallei Core Genome Multilocus Sequence Typing Scheme To Facilitate Molecular Surveillance. J Clin Microbiol. 2021;59(8):e0009321. pmid:33980649
- 12. Jolley KA, Bray JE, Maiden MCJ. Open-access bacterial population genomics: BIGSdb software, the PubMLST.org website and their applications. Wellcome Open Res. 2018;3:124. pmid:30345391
- 13. Trinh TT, Nguyen LDN, Nguyen TV, Tran CX, Le AV, Nguyen HV, et al. Melioidosis in Vietnam: Recently Improved Recognition but still an Uncertain Disease Burden after Almost a Century of Reporting. Trop Med Infect Dis. 2018;3(2):39. pmid:30274435
- 14. McCombie RL, Finkelstein RA, Woods DE. Multilocus sequence typing of historical Burkholderia pseudomallei isolates collected in Southeast Asia from 1964 to 1967 provides insight into the epidemiology of melioidosis. J Clin Microbiol. 2006;44(8):2951–62. pmid:16891516
- 15. Tran QTL, Phan PH, Bui LNH, Bui HTV, Hoang NTB, Tran DM, et al. Child Melioidosis Deaths Caused by Burkholderia pseudomallei-Contaminated Borehole Water, Vietnam, 2019. Emerg Infect Dis. 2022;28(8):1689–93. pmid:35697339
- 16. Trinh TT, Hoang TS, Tran DA, Trinh VT, Göhler A, Nguyen TT, et al. A simple laboratory algorithm for diagnosis of melioidosis in resource-constrained areas: a study from north-central Vietnam. Clin Microbiol Infect. 2018;24(1):84.e1-84.e4. pmid:28780059
- 17. Ashdown LR. An improved screening technique for isolation of Pseudomonas pseudomallei from clinical specimens. Pathology. 1979;11(2):293–7. pmid:460953
- 18. Novak RT, Glass MB, Gee JE, Gal D, Mayo MJ, Currie BJ, et al. Development and evaluation of a real-time PCR assay targeting the type III secretion system of Burkholderia pseudomallei. J Clin Microbiol. 2006;44(1):85–90. pmid:16390953
- 19. Hang NTT, Hang TTT, Trung TT. Initial investigation of pig with melioidosis caused by Burkholderia pseudomallei at Nghe An province. Vietnam Vet Assoc Vet Sci Tech. 2018;48–55.
- 20.
Hang NTT, Hang TTT, Trung TT, Khong NV. Initial investigation of Burkholderia pseudomallei in pigs in Nghe An province, Vietnam in 2016 and 2017. 9th World Melioidosis Conference. Hanoi, Vietnam; Oct. 15-18. 116 p.
- 21. Norris MH, Tran HTT, Walker MA, Bluhm AP, Zincke D, Trung TT, et al. Distribution of Serological Response to Burkholderia pseudomallei in Swine from Three Provinces of Vietnam. Int J Environ Res Public Health. 2020;17(14):5203. pmid:32708490
- 22. Trinh TT, Assig K, Tran QTL, Goehler A, Bui LNH, Wiede C, et al. Erythritol as a single carbon source improves cultural isolation of Burkholderia pseudomallei from rice paddy soils. PLoS Negl Trop Dis. 2019;13(10):e0007821. pmid:31634353
- 23. QGIS. QGIS Geographic Information System. 2024. Available from: http://qgis.org
- 24. Trung TT, Hetzer A, Göhler A, Topfstedt E, Wuthiekanun V, Limmathurotsakul D, et al. Highly sensitive direct detection and quantification of Burkholderia pseudomallei bacteria in environmental soil samples by using real-time PCR. Appl Environ Microbiol. 2011;77(18):6486–94. pmid:21803915
- 25. Norris MH, Zincke D, Daegling DJ, Krigbaum J, McGraw WS, Kirpich A, et al. Genomic and Phylogenetic Analysis of Bacillus cereus Biovar anthracis Isolated from Archival Bone Samples Reveals Earlier Natural History of the Pathogen. Pathogens. 2023;12(8):1065. pmid:37624025
- 26. Norris MH, Kirpich A, Bluhm AP, Zincke D, Hadfield T, Ponciano JM, et al. Convergent evolution of diverse Bacillus anthracis outbreak strains toward altered surface oligosaccharides that modulate anthrax pathogenesis. PLoS Biol. 2020;18(12):e3001052. pmid:33370274
- 27. Ewels PA, Peltzer A, Fillinger S, Patel H, Alneberg J, Wilm A, et al. The nf-core framework for community-curated bioinformatics pipelines. Nat Biotechnol. 2020;38(3):276–8. pmid:32055031
- 28. Peltzer A, Straub D, D VM, bot nf-core, Garcia MU, Taylor B, et al. nf-core/bacass: v2.1.0 nf-core/bacass: “Navy Steel Swordfish”. Zenodo; 2023.
- 29. Chen S, Zhou Y, Chen Y, Gu J. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 2018;34(17):i884–90. pmid:30423086
- 30. Wick RR, Judd LM, Gorrie CL, Holt KE. Unicycler: Resolving bacterial genome assemblies from short and long sequencing reads. PLoS Comput Biol. 2017;13(6):e1005595. pmid:28594827
- 31. Seemann T. Prokka: rapid prokaryotic genome annotation. Bioinformatics. 2014;30(14):2068–9. pmid:24642063
- 32. Silva M, Machado MP, Silva DN, Rossi M, Moran-Gilad J, Santos S, et al. chewBBACA: A complete suite for gene-by-gene schema creation and strain identification. Microb Genom. 2018;4(3):e000166. pmid:29543149
- 33. Horan K. kristyhoran/Coreugate. 2024. Available: https://github.com/kristyhoran/Coreugate
- 34. Shakya M, Ahmed SA, Davenport KW, Flynn MC, Lo C-C, Chain PSG. Standardized phylogenetic and molecular evolutionary analysis applied to species across the microbial tree of life. Sci Rep. 2020;10(1):1723. pmid:32015354
- 35. Nguyen L-T, Schmidt HA, von Haeseler A, Minh BQ. IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies. Mol Biol Evol. 2015;32(1):268–74. pmid:25371430
- 36. Zhou Z, Alikhan N-F, Sergeant MJ, Luhmann N, Vaz C, Francisco AP, et al. GrapeTree: visualization of core genomic relationships among 100,000 bacterial pathogens. Genome Res. 2018;28(9):1395–404. pmid:30049790
- 37. Letunic I, Bork P. Interactive Tree Of Life (iTOL) v4: recent updates and new developments. Nucleic Acids Res. 2019;47(W1):W256–9. pmid:30931475
- 38. Page AJ, Cummins CA, Hunt M, Wong VK, Reuter S, Holden MTG, et al. Roary: rapid large-scale prokaryote pan genome analysis. Bioinformatics. 2015;31(22):3691–3. pmid:26198102
- 39. Gee JE, Sacchi CT, Glass MB, De BK, Weyant RS, Levett PN, et al. Use of 16S rRNA gene sequencing for rapid identification and differentiation of Burkholderia pseudomallei and B. mallei. J Clin Microbiol. 2003;41(10):4647–54. pmid:14532197
- 40. Brennan S, Thompson JM, Gulvik CA, Paisie TK, Elrod MG, Gee JE, et al. Related Melioidosis Cases with Unknown Exposure Source, Georgia, USA, 1983-2024. Emerg Infect Dis. 2025;31(9):1802–6. pmid:40835221
- 41. Amemiya K, Dankmeyer JL, Fetterer DP, Worsham PL, Welkos SL, Cote CK. Comparison of the early host immune response to two widely diverse virulent strains of Burkholderia pseudomallei that cause acute or chronic infections in BALB/c mice. Microb Pathog. 2015;86:53–63. pmid:26162294
- 42. Welkos SL, Klimko CP, Kern SJ, Bearss JJ, Bozue JA, Bernhards RC, et al. Characterization of Burkholderia pseudomallei Strains Using a Murine Intraperitoneal Infection Model and In Vitro Macrophage Assays. PLoS One. 2015;10(4):e0124667. pmid:25909629
- 43. Depoorter E, De Canck E, Peeters C, Wieme AD, Cnockaert M, Zlosnik JEA, et al. Burkholderia cepacia Complex Taxon K: Where to Split?. Front Microbiol. 2020;11:1594. pmid:32760373
- 44. Hug S, Liu Y, Heiniger B, Bailly A, Ahrens CH, Eberl L, et al. Differential Expression of Paraburkholderia phymatum Type VI Secretion Systems (T6SS) Suggests a Role of T6SS-b in Early Symbiotic Interaction. Front Plant Sci. 2021;12:699590. pmid:34394152
- 45. Jeon J-S, Carreno-Quintero N, van Eekelen HDLM, De Vos RCH, Raaijmakers JM, Etalo DW. Impact of root-associated strains of three Paraburkholderia species on primary and secondary metabolism of Brassica oleracea. Sci Rep. 2021;11: 2781.
- 46. Stott CM, Bobay L-M. Impact of homologous recombination on core genome phylogenies. BMC Genomics. 2020;21(1):829. pmid:33238876
- 47. Norris MH, Heacock-Kang Y, Zarzycki-Siek J, Bluhm AP, McMillan IA, Schweizer HP, et al. Burkholderia pseudomallei natural competency and DNA catabolism: Identification and characterization of relevant genes from a constructed fosmid library. PLoS One. 2017;12(12):e0189018. pmid:29253888
- 48. Koonin EV, Rudd KE. Two domains of superfamily I helicases may exist as separate proteins. Protein Sci. 1996;5(1):178–80. pmid:8771213
- 49. Santos-Beneit F. The Pho regulon: a huge regulatory network in bacteria. Front Microbiol. 2015;6:402. pmid:25983732
- 50. Lamarche MG, Wanner BL, Crépin S, Harel J. The phosphate regulon and bacterial virulence: a regulatory network connecting phosphate homeostasis and pathogenesis. FEMS Microbiol Rev. 2008;32(3):461–73. pmid:18248418
- 51. Park J, Wang HH. Systematic dissection of σ70 sequence diversity and function in bacteria. Cell Rep. 2021;36(8):109590. pmid:34433066
- 52. Feklístov A, Sharon BD, Darst SA, Gross CA. Bacterial sigma factors: a historical, structural, and genomic perspective. Annu Rev Microbiol. 2014;68:357–76. pmid:25002089