Identification of a marker for two lineages within the GC1 clone of Acinetobacter baumannii

Sir, Isolates from a collection of 177 antibiotic-resistant Acinetobacter recovered at Westmead Hospital, Sydney over the period 1995– 2002 were further characterized. The relationships among these isolates had previously been examined by PFGE of ApaI-digested DNA, determination of MICs of several antibiotics and PCR screening to determine whether the blaOXA-23 gene or class 1 integrons were present. Eight PFGE pulsotypes were detected. The majority of isolates were A. baumannii but five isolates (PFGE pulsotypes III and IV) were not, and these were not examined here. One or two isolates from each year for each remaining pulsotype were tested using triplex PCRs targeting the oxa-Ab, csuE and ompA genes, to determine whether they belong to global clone 1 (GC1) or global clone 2 (GC2). Pulsotype I isolates were GC1 and isolates from pulsotypes VI, VII and VIII, which first appeared in 1999, were GC2. The 54 GC2 isolates were the only isolates resistant to imipenem and carried the blaOXA-23 gene. 1 Types II (n1⁄464) and V (n1⁄42) did not belong to either global clone. This is consistent with a previous report using a smaller number of isolates. Pulsotype I included 52 isolates and was present in the hospital intensive care unit from 1995 to 1999, with a single isolate recovered in 2000. Isolates in this group are among the earliest multiply antibiotic-resistant A. baumannii reported in Australia. They were gentamicin resistant and carried the aacC1-orfP-orfP-orfQ-aadA1 cassette array in a class 1 integron. They were also resistant to ciprofloxacin but susceptible to third-generation cephalosporins and carbapenems. Consistent with the GC1 designation, WM98, one of two representatives examined previously, was found here to belong to sequence type ST109 [Oxford multilocus sequence typing (MLST) scheme]. Ten representative isolates, including WM98, covering each of the years 1995–99, were examined in more detail, and, in contrast to the previous report, WM98 and other members of this group did not include a copy of ISAba1. The resistance profile was expanded using disc diffusion, and the resistance genes present were determined by PCR. The comM gene was interrupted, and they were found to also carry the sul1 sulphonamide resistance gene, the tet(A) tetracycline resistance determinant and transposon Tn6020 carrying the aphA1b kanamycin and neomycin resistance gene. The catA1 and blaTEM genes were also present. PCR mapping 5 of WM98 together with sequencing of all IS26 junctions revealed the presence of a genomic resistance island very closely resembling AbaR3 in AB0057 (GenBank accession number CP001182). We had noticed that AbaR3 contained a deletion of 108 bp in the 5′-conserved segment (5′-CS) of the class 1 integron, and that this deletion was not present in AbaR5, which now has been completely sequenced (GenBank accession number FJ172370), or in AbaR1 in AYE (GenBank accession number CU459141) or in AbaR2 in the GC2 isolate ACICU (GenBank accession number CP000863). The location of this deletion is indicated in Figure 1(a). It removes the last 52 bp of the intI1 gene, and the loss of conserved amino acids Research letters