ABSTRACT
UNLABELLED: Whole-genome sequence (WGS) data are commonly used to design diagnostic targets for the identification of bacterial pathogens. To do this effectively, genomics databases must be comprehensive to identify the strict core genome that is specific to the target pathogen. As additional genomes are analyzed, the core genome size is reduced and there is erosion of the target-specific regions due to commonality with related species, potentially resulting in the identification of false positives and/or false negatives. IMPORTANCE: A comparative analysis of 1,130 Burkholderia genomes identified unique markers for many named species, including the human pathogens B. pseudomallei and B. mallei Due to core genome reduction and signature erosion, only 38 targets specific to B. pseudomallei/mallei were identified. By using only public genomes, a larger number of markers were identified, due to undersampling, and this larger number represents the potential for false positives. This analysis has implications for the design of diagnostics for other species where the genomic space of the target and/or closely related species is not well defined.
Subject(s)
Burkholderia/isolation & purification , Genome, Bacterial , Sequence Analysis, DNA , Bacterial Typing Techniques , Burkholderia/classification , Burkholderia/genetics , Databases, Genetic , False Positive Reactions , Genetic Markers , Humans , Pathology, Molecular/methodsABSTRACT
The global distribution of the soil-dwelling bacterium Burkholderia pseudomallei, causative agent of melioidosis, is poorly understood. We used established culturing methods developed for B. pseudomallei to isolate Burkholderia species from soil collected at 18 sampling sites in three states in the southern United States (Arizona (n = 4), Florida (n = 7), and Louisiana (n = 7)). Using multi-locus sequence typing (MLST) of seven genes, we identified 35 Burkholderia isolates from these soil samples. All species belonged to the B. cepacia complex (Bcc), including B. cenocepacia, B. cepacia, B. contaminans, B. diffusa, B. metallica, B. seminalis, B. vietnamiensis and two unnamed members of the Bcc. The MLST analysis provided a high level of resolution among and within these species. Despite previous clinical cases within the U.S. involving B. pseudomallei and its close phylogenetic relatives, we did not isolate any of these taxa. The Bcc contains a number of opportunistic pathogens that cause infections in cystic fibrosis patients. Interestingly, we found that B. vietnamiensis was present in soil from all three states, suggesting it may be a common component in southern U.S. soils. Most of the Burkholderia isolates collected in this study were from Florida (30/35; 86%), which may be due to the combination of relatively moist, sandy, and acidic soils found there compared to the other two states. We also investigated one MLST gene, recA, for its ability to identify species within Burkholderia. A 365bp fragment of recA recovered nearly the same species-level identification as MLST, thus demonstrating its cost effective utility when conducting environmental surveys for Burkholderia. Although we did not find B. pseudomallei, our findings document that other diverse Burkholderia species are present in soils in the southern United States.