Genomic Diversity of Burkholderia pseudomallei in Ceara, Brazil

Abstract

Burkholderia pseudomallei is a Gram-negative bacterium that causes the sapronotic disease melioidosis. An outbreak in 2003 in the state of Ceara, Brazil, resulted in subsequent surveillance and environmental sampling which led to the recognition of B. pseudomallei as an endemic pathogen in that area. From 2003 to 2015, 24 clinical and 12 environmental isolates were collected across Ceara along with one from the state of Alagoas. Using next-generation sequencing, multilocus sequence typing, and single nucleotide polymorphism analysis, we characterized the genomic diversity of this collection to better understand the population structure of B. pseudomallei associated with Ceara. We found that the isolates in this collection form a distinct subclade compared to other examples from the Western Hemisphere. Substantial genetic diversity among the clinical and environmental isolates was observed, with 14 sequence types (STs) identified among the 37 isolates. Of the 31,594 core single-nucleotide polymorphisms (SNPs) identified, a high proportion (59%) were due to recombination. Because recombination events do not follow a molecular clock, the observation of high occurrence underscores the importance of identifying and removing recombination SNPs prior to evolutionary reconstructions and inferences in public health responses to B. pseudomallei outbreaks. Our results suggest long-term B. pseudomallei prevalence in this recently recognized region of melioidosis endemicity.IMPORTANCEB. pseudomallei causes significant morbidity and mortality, but its geographic prevalence and genetic diversity are not well characterized, especially in the Western Hemisphere. A better understanding of the genetic relationships among clinical and environmental isolates will improve knowledge of the population structure of this bacterium as well as the ability to conduct epidemiological investigations of cases of melioidosis.

Keywords: environmental; genome analysis; infectious disease; melioidosis; molecular epidemiology.

FIG 1

Maximum-likelihood phylogenetic tree of selected B. pseudomallei isolates collected throughout the world. Each leaf is labeled with its strain (or isolate) designation as the prefix, and geographic locations of isolation origin are suffixes. Bootstrap values above 70% are shown for more internal nodes, but none are shown for those near leaves. The term “ex” indicates the putative location where infection occurred. N.E., northeastern. The substitution bar indicates the number of substitutions per SNP site.

FIG 2

Maximum-likelihood phylogenetic tree of genome assemblies from northeast Brazil. Core SNP positions (n = 31,594) for each sample and each reconstructed ancestral node were labeled as mutational (light blue) or due to recombination (dark blue) on a per-site basis. Dotted horizontal lines connect samples and internal ancestral sequences to the illustration of inferred recombination positions. Cluster membership is indicated with color-coded circles and their corresponding-colored leaves, which contain MLST assignments for comparison. Isolate quantities per cluster are noted in parentheses in the key.

FIG 3

Geographic distribution of SNP-based cluster types associated with isolates from across the state of Ceara. Each circle represents one genome of a given cluster and is placed adjacent to the name of the municipality with which it is associated. The dots do not directly correspond to a specific map coordinate. Three isolates are not shown: one from the state of Alagoas and two others known to be from Ceara but not associated with a specific municipality.