Infections caused by extensively-antibiotic resistant (XAR) Acinetobacter baumannii belonging to global clone 2 (GC2) are a major problem in hospitals worldwide. To understand the evolution and epidemiology of this organism in Australia, 91 GC2 A. baumannii collected between 1999-2011 from 8 hospitals in Sydney, Brisbane and Canberra were investigated. The structures of genomic islands and plasmids were studied to determine if they could be used as markers to track spread.
Draft genomes generated using Illumina HiSeq and assembled using Velvet were searched for the components of genomic islands and plasmids. These were assembled using known structures as scaffolds and predictions were confirmed using PCR and sequencing. MLST profiles revealed that every isolate except one was ST208. Every isolate also shared a common genomic island, AbGRI1, that contained oxa23 in Tn2006, sul2, tetA(B) and strA-strB. In Australian isolates, AbGRI1 harbours Tn2006 in a characteristic location as well as a rare IS, ISAba17. To date, this combination has not been reported in GC2 isolates outside Australia.
However, the collection had several aminoglycoside resistance phenotypes and genes. Two of the aminoglycoside resistance genes were located together in the chromosome in a second genomic island, AbGRI2. Variation within this island due to IS26-mediated deletions has removed segments with resistance genes, causing some of the phenotypic variation. In some cases, aminoglycoside resistance was restored or extended by plasmid acquisition. Specific AbGRI2 variants and plasmids were found to correspond with different outbreaks within a hospital, while others were seen in isolates from hospitals in more than one city, indicating transmission.
This collection of isolates clearly stemmed from a common ancestor, indicative of a single entry into Australia. Subsequent changes in resistance plasmids, cryptic plasmids and AbGRI2 have led to divergence within this otherwise closely related GC2 A. baumannii population.