Increased knowledge of Francisella genus diversity highlights the benefits of optimised DNA-based assays
- Equal contributors
1 Division of CBRN Security and Defence, FOI, Swedish Defence Research Agency, SE- 906 21, Umeå, Sweden
2 Department of Clinical Microbiology, Umeå University, SE–901 85, Umeå, Sweden
3 Laboratory for Molecular Infection Medicine Sweden (MIMS), Umeå University, SE–901 87, Umeå, Sweden
BMC Microbiology 2012, 12:220 doi:10.1186/1471-2180-12-220Published: 25 September 2012
Recent advances in sequencing technologies offer promising tools for generating large numbers of genomes, larger typing databases and improved mapping of environmental bacterial diversity. However, DNA-based methods for the detection of Francisella were developed with limited knowledge about genetic diversity. This, together with the high sequence identity between several Francisella species, means there is a high risk of false identification and detection of the highly virulent pathogen Francisella tularensis. Moreover, phylogenetic reconstructions using single or limited numbers of marker sequences often result in incorrect tree topologies and inferred evolutionary distances. The recent growth in publicly accessible whole-genome sequences now allows evaluation of published genetic markers to determine optimal combinations of markers that minimise both time and laboratory costs.
In the present study, we evaluated 38 previously published DNA markers and the corresponding PCR primers against 42 genomes representing the currently known diversity of the genus Francisella. The results highlight that PCR assays for Francisella tularensis are often complicated by low specificity, resulting in a high probability of false positives. A method to select a set of one to seven markers for obtaining optimal phylogenetic resolution or diagnostic accuracy is presented.
Current multiple-locus sequence-typing systems and detection assays of Francisella, could be improved by redesigning some of the primers and reselecting typing markers. The use of only a few optimally selected sequence-typing markers allows construction of phylogenetic topologies with almost the same accuracy as topologies based on whole-genome sequences.