, 2007a, b) and compared to a larger this website published database of P. aeruginosa isolates from nonocular sources (Stewart et al., 2011). Various markers in the set of P. aeruginosa isolates associated with keratitis were discordant with the wider P. aeruginosa population. These included previously reported associations, such as carriage of exoU. It was also demonstrated that 17 of 63 (27%) keratitis isolates from 2003 to 2004 carried a distinctive allele of pilA, the gene that encodes the pilin of type IV pili. Thus, the keratitis isolates were associated

with specific characteristics, suggesting that a subpopulation of P. aeruginosa may be adapted to causing corneal infections (Stewart et al., 2011). However, we could not be sure whether these genetic features of keratitis-associated isolates would be consistent temporally or represented a feature of the particular time Cisplatin period chosen

for sampling. To address this question, in this study, we report the analysis of a set of keratitis-associated P. aeruginosa isolates, collected by the MOG from patients in the UK, during a different time period, 5 years later. Sixty isolates (listed in Table 1) from corneal scrape samples were collected from patients with bacterial keratitis (2009–2010) from the six hospitals comprising the MOG. DNA was purified using the Wizard Genomic DNA purification kit (Promega, UK), as per the manufacturer’s instructions. A further 18 isolates of P. aeruginosa from bloodstream infections (collected and stored in Liverpool 2010–2011) Fludarabine were also used. All isolates tested positive for the oprL gene using a P. aeruginosa-specific PCR assay (De Vos et al., 1997). Genotyping of P. aeruginosa was conducted using the AT genotyping system (Wiehlmann et al., 2007a, b; Alere Technologies, Jena, Germany), as per the manufacturer’s instructions. Analysis of 13 single nucleotide polymorphisms (SNPs) based on the conserved genome, and three variable markers (flagellin types a or b and the mutually exclusive type III secretion exotoxin genes exoU or exoS),

was used to generate a four character hexadecimal code as described previously (Wiehlmann et al., 2007a, b; Stewart et al., 2011). This hexadecimal code was used to assign specific clone types. The genotypic relationships between keratitis isolates of P. aeruginosa and nonocular isolates were assessed by analysing each strain for 14 binary markers as described previously (Stewart et al., 2011). Presence or absence of exoS or exoU was not included in this analysis. The wider population was represented by a database of 322 nonkeratitis P. aeruginosa isolates, representing 128 clones, taken from various sources (Wiehlmann et al.,2007a, b; Mainz et al., 2009; Rakhimova et al., 2009). The analysis was undertaken using the eBURST(v3) algorithm (Feil et al., 2004; Spratt et al., 2004).