The complete sequence of Salmonella genomic island SGI2

Sir, Complete structure of SGI2 and variants. (a) The structure of SGI2 is drawn to scale from GenBank accession number AY463797 with the class 1 integron containing the resistance genes below the backbone. (b) Variant forms of the SGI2 integron are shown. Genes and ORFs are shown below the line as arrows indicating the direction of transcription and named below with ‘5’ and ‘6’ representing orf5 and orf6. The chromosomal segments flanking the backbone are represented by thicker lines; the trmE gene was previously known as thdF. In the integron, the gene cassettes are narrow open boxes with a bar representing the attC site and an open bar representing attI1. ISs are shown as boxes with the IS number inside. The segment surrounding CR3 that is not a normal constituent of class 1 integrons is represented as an open box of medium width. SGI2 was first found in the S. enterica serovar Emek strain SRC19, which was recovered from sewage effluent in September 1999 in Australia and was resistant to chloramphenicol, sulfamethoxazole, tetracycline and trimethoprim as well as nalidixic acid.4 We originally reported SGI2 as a variant of SGI1 (SGI1-J).2 However, further analysis revealed that it could not have arisen from SGI1 and hence was not an SGI1 variant.4 We therefore renamed this island SGI2.2 The structure of the complex class 1 integron was later revised to include an additional partial tni module and part of the backbone was also sequenced, revealing a significant number of single nucleotide differences between the SGI1 and SGI2 backbones.3 SGI2 was also found in serovar Emek isolates from the UK (1999) and a traveller returning to Australia from Thailand in 2002.4 SGI2-A, a deletion derivative that has lost the cmlA9 gene (chloramphenicol and florfenicol resistance4) and tet(G) determinant (tetracycline resistance), was also found in a returning traveller.4 In order to facilitate future work on SGI2 and to compare the complete SGI2 backbone with that of SGI1, here we have completed the sequence of SGI2 by sequencing the genome of SRC19 on an Illumina HiSeq at the Australian Genomic Research Facility. The reads (30-fold average read depth) were assembled using Velvet,5 contigs containing parts of SGI2 were recovered and the complete sequence of SGI2 was assembled using our available sequence for the complex class 1 integron and part of the backbone (GenBank accession number AY963803)3 as a template. In the 27 364 bp backbone there are 144 single base differences from the SGI1 backbone, representing 0.53% difference between the two backbones. However, these differences are not evenly distributed along the backbone and with clusters removed the difference is 0.16%. Using divergence estimates for Salmonella,6 this corresponds to separation of the two lineages 2000–8000 years ago. The ST of SRC19 determined from the genome sequence data was ST76 (aroC8, dnaN38, hemD40, hisD12, purE5, sucA9, thrA38). SRC19 is resistant to nalidixic acid (7 mm diameter zone with a disc containing 30 μg of nalidixic acid), but has only reduced susceptibility to ciprofloxacin (22 mm diameter zone with a disc containing 5 μg of ciprofloxacin). Consistent with this phenotype, codon 83 of the gyrA gene is TAC encoding a tyrosine residue, which would usually be associated with nalidixic acid resistance, whereas codon 80 of parC is AGC encoding a serine residue, which would usually be associated with susceptibility to ciprofloxacin. The SGI2-carrying serovar Emek strain is believed to have originated in Asia.1 However, like SGI1, SGI2 has spread among S. enterica serovars. SGI2 (named SGI1-J3; GenBank accession number EU924797) has been found in serovar Virchow isolated in Taiwan in 1993 and 1994.7 Later serovar Virchow isolates from Taiwan (2004–06) examined using PCR and hybridization also carried SGI2 or a variant of it (called SGI1-J4 in Chu et al.,8 but designated SGI2-B in Figure 1) that had lost all of the antibiotic resistance genes except sul1 (sulphonamide resistance).8 Recently, S. enterica serovar Kentucky ST198 isolates have been reported to include SGI2-B or a further variant (called SGI1-J6 in Le Hello et al.,9 but SGI2-C in Figure 1) that can be derived from SGI2 by replacement of the dfrA1-orfC cassettes with the aadA2 cassette and incorporation of an IS.9 Until recently, SGI1, found in S. enterica serovar Typhimurium DT104, was the only completely sequenced example of this island group (GenBank accession number AF261825). The sequence reported here provides the complete sequence of SGI2 and this will facilitate its detection in Salmonella genomes and form a sound basis for assembly of SGI2 and SGI2 variant forms. As these variant forms clearly arise from SGI2 in a similar manner to the variant forms of SGI1, assignment of names using a system of nomenclature parallel to that used for SGI1 is clearly preferable. The sequence reported here was added to GenBank accession number AY963803. This study and M. H. were supported by NHMRC Project Grant APP1060608. K. E. H. is supported by an NHMRC Career Development Fellowship (APP10610409). None to declare.