Validation of partial rpoB gene sequence analysis for the identification of clinically important and emerging Acinetobacter species

Abstract

Bacteria belonging to the genus Acinetobacter are ubiquitous in soil and water. Only a few species, including Acinetobacter baumannii, and the unnamed Acinetobacter genomic species (gen. sp.) 3 and 13TU, which together with the soil organism Acinetobacter calcoaceticus are combined in the A. calcoaceticus-A. baumannii (Acb) complex, have been recognized as important nosocomial infectious agents. The ecology, epidemiology and pathology of most species are not yet well established. Lack of practical and accurate methods limits routine identification of clinical isolates and thus hampers precise identification of those of the Acb complex and other Acinetobacter species of possible clinical significance. We previously identified a 350 bp highly variable zone on the rpoB gene which appeared to be a promising target for rapid molecular identification. In the present study, we validated this method for accuracy on a collection of reference strains belonging to A. calcoaceticus (5 strains), Acinetobacter gen. sp. 3 (29 strains), A. gen. sp. 13TU (18 strains), A. baumannii (30 strains) and one strain each of A. radioresistens, A. gen. sp. 15TU, A. gen. sp. 10, A. gen. sp. 11, A. gen. sp. 'between 1 and 3' and A. gen. sp. 14TU=13BJ. This represents the largest analysis to date that compares a large number of well-identified strains of the Acb complex to assess the intra- and interspecies variation within this complex. All were correctly identified with 98.9-100 % intraspecies relatedness based on partial rpoB sequence analysis. We then applied this tool to identify 99 Acinetobacter clinical isolates from four public hospitals in Marseille, France. All isolates could easily be identified to species as they were separated into 13 species sequence types with a sequence variance of 0-2.6% from their respective type strains. Of these 99 isolates, 10 were A. haemolyticus, 52 were A. baumannii, 27 were A. gen. sp. 3, 5 were A. schindleri, 1 was A. lwoffii, and 1 was A. gen. sp. 13TU. Three were provisionally identified as A. gen. sp. 9. This is the first work to identify all specimens of a set of clinical Acinetobacter isolates at species level using rpoB sequence analysis. Our data emphasize the recognition of A. schindleri as an emerging cause of Acinetobacter-related infection and confirm that A. gen. sp. 3 is the second most commonly isolated Acinetobacter species after A. baumannii in patients.