Genotypic Diversity within a Single Pseudomonas aeruginosa Strain Commonly Shared by Australian Patients with Cystic Fibrosis

In cystic fibrosis (CF), Pseudomonas aeruginosa undergoes intra-strain genotypic and phenotypic diversification while establishing and maintaining chronic lung infections. As the clinical significance of these changes is uncertain, we investigated intra-strain diversity in commonly shared strains from CF patients to determine if specific gene mutations were associated with increased antibiotic resistance and worse clinical outcomes. Two-hundred-and-one P. aeruginosa isolates (163 represented a dominant Australian shared strain, AUST-02) from two Queensland CF centres over two distinct time-periods (2001–2002 and 2007–2009) underwent mexZ and lasR sequencing. Broth microdilution antibiotic susceptibility testing in a subset of isolates was also performed. We identified a novel AUST-02 subtype (M3L7) in adults attending a single Queensland CF centre. This M3L7 subtype was multi-drug resistant and had significantly higher antibiotic minimum inhibitory concentrations than other AUST-02 subtypes. Prospective molecular surveillance using polymerase chain reaction assays determined the prevalence of the ‘M3L7’ subtype at this centre during 2007–2009 (170 patients) and 2011 (173 patients). Three-year clinical outcomes of patients harbouring different strains and subtypes were compared. MexZ and LasR sequences from AUST-02 isolates were more likely in 2007–2009 than 2001–2002 to exhibit mutations (mexZ: odds ratio (OR) = 3.8; 95% confidence interval (CI): 1.1–13.5 and LasR: OR = 2.5; 95%CI: 1.3–5.0). Surveillance at the adult centre in 2007–2009 identified M3L7 in 28/509 (5.5%) P. aeruginosa isolates from 13/170 (7.6%) patients. A repeat survey in 2011 identified M3L7 in 21/519 (4.0%) P. aeruginosa isolates from 11/173 (6.4%) patients. The M3L7 subtype was associated with greater intravenous antibiotic and hospitalisation requirements, and a higher 3-year risk of death/lung transplantation, than other AUST-02 subtypes (adjusted hazard ratio [HR] = 9.4; 95%CI: 2.2–39.2) and non-AUST-02 strains (adjusted HR = 4.8; 95%CI: 1.4–16.2). This suggests ongoing microevolution of the shared CF strain, AUST-02, was associated with an emerging multi-drug resistant subtype and possibly poorer clinical outcomes.


Introduction
Initial pulmonary infections with Pseudomonas aeruginosa in patients with cystic fibrosis (CF) are typically intermittent and caused by non-mucoid environmental strains that are susceptible to anti-pseudomonal antibiotics [1][2][3]. Eventually, a single strain becomes established and develops a mucoid phenotype, which with increasing antibiotic resistance makes eradication from the airways difficult [4]. Once this event occurs, patients experience an accelerated pulmonary decline with increased exacerbation frequency and mortality [5].
While environmental origins of P. aeruginosa in CF patients are well recognised, reports of patients sharing indistinguishable strains suggest cross-infection as another acquisition pathway [6,7]. Importantly, shared CF strains can have different phenotypic properties and clinical impact. In 2004 the Liverpool Epidemic strain (LES) accounted for 11% of P. aeruginosa infected patients attending 15 surveyed CF clinics in England and Wales [8]. Compared with other P. aeruginosa strains, it had increased antibiotic resistance and was associated with greater declines in lung function and nutritional status [9,10]. LES was also found in 15% of patients attending CF clinics in Ontario, Canada where it was associated with an increased risk of death or lung transplantation [11]. Similarly, the Manchester Epidemic Strain, detected in 14% of patients attending a Manchester CF clinic was multi-drug resistant (MDR) and associated with increased antibiotic and healthcare use, but not mortality [12]. In Australia, a national microbiological survey of CF centres found two strains, AUST-01 and AUST-02, in 22% and 18% of CF patients harbouring P. aeruginosa respectively [13]. Furthermore, regional differences existed and AUST-02 was observed in 35%-40% of infected patients attending Queensland CF clinics [13,14]. This strain had increased carbapenem and aminoglycoside antibiotic resistance and was associated with greater treatment requirements than unique strains [13,15]. However, the actual long-term clinical impact of AUST-02 remains uncertain.
In addition to between-strain differences, extensive intra-strain diversity is observed in P. aeruginosa isolates from chronically infected CF airways [16][17][18][19][20]. The CF lung is a spatially heterogeneous structure resulting in differential exposures to antibiotics, host defences, oxidative stress, nutrition and oxygen availability. Consequently, the resident P. aeruginosa population demonstrates extensive in-vivo intra-strain genotypic and phenotypic diversity (or adaptive radiation) involving antibiotic susceptibility, colonial morphology, motility, biofilm formation, auxotrophy, quorum sensing and virulence factor expression during the course of chronic infection [16][17][18][19][20]. Genes encoding key regulatory networks, as well as metabolism, antibiotic resistance and virulence factors are particularly prone to mutations during chronic CF airway infection and the key regulatory genes for multi-drug efflux pumps (e.g. mexZ) and virulence factors (e.g. lasR) are among the most common mutational targets [16,21,22] mexZ inhibits the MexXY-OprM efflux pump, which is over-expressed when this gene is inactivated resulting in P. aeruginosa becoming less susceptible to aminoglycoside, β-lactam and fluoroquinolone antibiotics [23]. In contrast, inactivation of lasR decreases the expression of several virulence genes and confers a growth advantage in the amino acid enriched mucous secretions  [24], while harbouring P. aeruginosa with lasR mutations is associated with accelerated progression of CF lung disease [25]. Nevertheless, whether these mutations affect transmissibility, morbidity and mortality of shared P. aeruginosa strains remains unclear. In this study, we used mexZ and lasR DNA sequencing to: (i) investigate intra-strain diversity within P. aeruginosa isolates from Queensland CF patients (involving mostly the commonly shared AUST-02 strain) and (ii) determine if mutations in these genes are associated with increased antibiotic resistance and poorer clinical outcomes.

Patient isolates and setting
This study, outlined in Figs 1 and 2, utilised stored P. aeruginosa isolates from two previous cross-sectional surveys [13,15,26], as well as the ongoing longitudinal Australian Clonal P. aeruginosa in Cystic Fibrosis (ACPinCF) study [13], which is linked with the Australian CF Data Registry (ACFDR To assess intra-strain diversity within the mexZ and lasR gene sequences of P. aeruginosa strains, 201 stored P. aeruginosa isolates were assessed for mexZ and lasR sequence diversity. These isolates had been collected previously from 107 Queensland patients with CF attending either the Royal Children's Hospital (RCH) or The Prince Charles Hospital (TPCH), Brisbane, Australia and were part of microbiological surveys conducted in 2001-2002 and 2007-2009 respectively (Fig 1) [13,15,26]. The isolates originated from sputum samples and had already been genotyped by enterobacterial repetitive intergenic consensus-polymerase chain reaction (ERIC-PCR) and Sequenom iPlex single nucleotide polymorphism (SNP)-based methods in earlier pointprevalence and typing studies [13,26]. They included 163 AUST-02 isolates and 38 other minor shared and unique strains (S1 Table). mexZ and lasR gene PCR assays were performed using primers displayed in the S2 Table. Automated DNA sequencing of the amplicons was undertaken by The Australian Genome Research Facility using Big Dye Terminator chemistry version 3.1 (Applied Biosystems) and an Applied Biosystems 3730 capillary sequencer. Sequencing data were used to assign a subtype based on a mexZ intergenic and partial coding nucleotide sequence (approximately 755 bases) and the complete LasR protein (240 amino acids). As outlined in Fig 1, the minimum inhibitory concentrations (MIC) for a subset of isolates to meropenem, imipenem, ceftazidime, ticarcillin-clavulanic acid, aztreonam, ciprofloxacin, tobramycin and colistin sulphate were determined by broth microdilution testing following the Clinical and Laboratory Standards Institute (CLSI) 2013 guidelines [27]. MDR was defined as resistance to all agents tested within two or more antibiotic groups, including the aminoglycosides, β-lactam agents and the fluoroquinolone, ciprofloxacin, according to the CF Foundation consensus guidelines [28,29].  component of the ACPinCF study [13], patients were already providing one-to-three P. aeruginosa isolates from single sputum specimens obtained annually for culture. It was therefore possible to screen systematically for this M3L7 subtype in 509 P. aeruginosa isolates prospectively collected from the 170 infected adult CF patients attending this clinic between 2007-2009 and another 519 isolates from 173 patients attending the same clinic during 2011 (Fig 2). 'M3L7' detection was by SYBR-green based allele-specific real-time PCR assay (M3-PCR) targeting a SNP specific to the 'M3' mexZ sequence (primers detailed in the S2 Table). Isolates positive by M3-PCR had lasR DNA sequencing to confirm an M3L7 genotype. These M3L7 isolates underwent strain typing using a validated AUST-02 allele specific PCR assay (S2 Table).

Clinical data collection
To assess the clinical impact of the M3L7 subtype, patients with P. aeruginosa identified in the 2007-2009 microbiological survey were followed longitudinally for up to 4-years from the time of their baseline sample collection until the census date of January 1 st , 2011. Baseline clinical variables, including age, gender, CF genotype, pancreatic insufficiency, CF-related diabetes and/or liver disease, body-mass-index (BMI) at the time of the recorded forced expiratory volume in the first second (FEV 1 [% predicted]) in the calendar year of sputum collection, the aforementioned FEV 1 % predicted, co-pathogens (Burkholderia cepacia complex, Staphylococcus aureus, Achromobacter and Aspergillus species), maintenance treatment (azithromycin, inhaled antibiotics and/or mucolytic therapies), as well as treatment episodes requiring intravenous (IV) antibiotics (number of episodes and days of hospitalisation) and the number of outpatient visits in the 12-months before sputum collection were recorded as described previously [13]. Longitudinal clinical data from individual patients were retrieved from the ACFDR for BMI, lung function measurements, frequency of hospitalisations and outpatient clinic visits, from the time of baseline isolate collection to the end of the study, death or lung transplantation, whichever came sooner.

Statistical analyses
Clinical variables and MIC results were summarised using descriptive statistics. Continuous variables were expressed as either mean (standard deviation, SD) or median (interquartile range, IQR) values. Normally distributed data were compared using paired t-test (2 groups) and analysis of variance (>2 groups) with post-hoc Bonferroni tests. Continuous data that were not normally distributed were analysed by either the Mann-Whitney-U test (2 group comparison) or the Kruskal-Wallis test (>2 groups) with post-hoc Bonferroni tests. Categorical data were expressed as percentages and compared by Chi-square and Fishers' Exact tests. Pvalues were 2-sided, with odds ratios (OR) and 95% confidence intervals (CI). All these analyses were performed using the SPSS software version 22.
To assess for a differential clinical impact by various P. aeruginosa strains or subtypes, patients were categorised into one of three groups according to their P. aeruginosa typing results: (i) M3L7, (ii) non-M3L7 AUST-02 subtypes and (iii) non AUST-02 strains. Patients co-infected with both M3L7 and non M3L7 strains were classified as having M3L7. The primary outcome was death or lung transplantation from CF-related respiratory failure, while all other patients were censored (defined as alive on January 1 st , 2011, death due to causes other than end-stage pulmonary disease or transfer from the centre). Survival time was measured in days from sample collection until endpoint events or censoring. Kaplan-Meier survival curves of unadjusted time to death or lung transplantation were generated for the three patient groups. Log-rank test was performed to assess between-group differences in survival curves. Adjustment for confounding factors was performed using univariate and multivariate Cox proportional hazards regression models. Potential confounding factors were selected a priori and included age, gender, CF genotype, pancreatic insufficiency, CF-related diabetes and/or liver disease, BMI, severity of airflow obstruction (categorised as mild, FEV 1 >70%; moderate, FEV 1 40%-70%; severe, FEV 1 <40%) and co-pathogens (B. cepacia complex, S. aureus, Achromobacter and Aspergillus sp.). The effect of each factor was assessed by univariate Cox-proportional analysis and factors identified as statistically significant (P-value 0.05) were included in the multivariate analysis. Multivariate analysis based on the Cox proportional hazard regression model was performed using a stepwise backward method based on likelihood ratios using P0.05 for inclusion. The proportional hazards assumption of the model was satisfied by model diagnostics including: log minus log-survival curves, diagnostic plots of Dfbeta, and estimates of martingale residues and deviance residuals. The decision to incorporate a covariate as a categorical or continuous variable was based on model diagnostic test results.
To compare the rates of disease progression between the three patient groups, multivariate repeated measures linear regression modelling was employed to account for within-patient correlation and to estimate the annual rates of change in BMI and FEV 1 (% predicted) after adjusting for gender effect. Dunn's test and post-hoc Bonferroni tests were then performed to compare rates of change in clinical variables between patient groups.

Combined mexZ and LasR subtypes
Sixty-two mexZ/LasR sequence combinations (referred to as 'subtypes'; S1 Table) were observed. Of the 150 sequenced AUST-02 isolates (79 patients), 35 different mexZ/LasR subtypes were identified. The M2L1 subtype was detected most commonly (72/150 isolates, 48%) and was considered the wild-type AUST-02 subtype, as it included the wild-type L1 sequence and was the most prevalent AUST-02 subtype in the early study years. Of the other 34 AUST-02 subtypes, 28 were from individuals and six were identified in clusters of two or more patients. The latter included M2L3 (four unrelated patients), M2L20 (two siblings and one unrelated patient), M2L26 (two siblings), M2L28 (two unrelated patients), M2L30 (two unrelated patients) and M3L7 (10 unrelated patients). Whilst most of these AUST-02 subtypes were confined to small patient clusters (median: 2.5 patients/cluster), 16 isolates identified as the M3L7 AUST-02 subtype were found in 10 unrelated patients whose care was based at TPCH in 2007-2009. This M3L7 AUST-02 subtype was not detected in paediatric patients attending the RCH or from the earlier collection period (2001)(2002) at either centre.

Antibiotic susceptibility
Antibiotic susceptibilities to several anti-pseudomonal antibiotics were then assessed for the 16 M3L7 isolates and in a subset of 58 isolates randomly selected from across all non-M3L7 AUST-02 subtypes (S3 Table). All 16 M3L7 isolates were MDR and demonstrated non-susceptibility (intermediate or resistant) to a significantly greater number of antibiotics than non-M3L7 AUST-02 isolates (Table 1). Except for tobramycin, M3L7 isolates also had significantly higher MIC values for all other antibiotics tested than non-M3L7 AUST-02 subtypes. In 30 patients, more than one P. aeruginosa isolate was sampled from the same sputum specimen and each of these isolates underwent mexZ and lasR sequencing. Twenty patients had isolates of the same P. aeruginosa strain with identical mexZ/LasR subtypes; three had different P. aeruginosa strains, and seven had two to three different AUST-02 mexZ/LasR subtypes within the same sputum sample. Forty patients also had serial isolates collected at both time-points (2001-2002 and 2007-2009; Table 2A). Alterations in P. aeruginosa strains and/or intra-strain mexZ/LasR subtypes were seen in 17 (42.5%) of these patients (Table 2A), including two with new P. aeruginosa strains, two with changes in both P. aeruginosa strains and mexZ/LasR subtypes, and another 13 with changes in AUST-02 mexZ/LasR subtypes.

Point-prevalence studies of the 'M3L7' AUST-02 subtype in 2007-2009 and 2011
Following the initial identification of the 'M3L7' AUST-02 subtype, we took advantage of cross-sectional microbiological surveys [13] of all known P. aeruginosa infected CF adult patients at TPCH that were conducted during 2007-2009 and repeated in 2011. Stored isolates from these two surveys were tested by PCR assays designed specifically to detect the M3L7 subtype within the AUST-02 strains (Fig 2).

Baseline Clinical Characteristics
Of the 170 patients surveyed in 2007-2009, 166 were available for follow-up and four were excluded as the strain types of their baseline isolates were unavailable (Fig 3). Table 3 summarises their baseline characteristics with patients categorised according to whether they had: (i) M3L7, (ii) non-M3L7 AUST-02 subtypes or (iii) non AUST-02 P. aeruginosa strains. Whilst the mean baseline FEV 1 (% predicted) was lower in the M3L7 group (45.5%) compared with both the non M3L7 AUST-02 group (54.7%) and the non AUST-02 groups (52.8%), the differences did not reach statistical significance. However, patients with M3L7 had lower BMI and in the previous 12-months had undertaken more frequent outpatient visits than patients with   non AUST-02 strains. Patients with M3L7 had also received more hospital-based IV antibiotic treatments than patients with other non-M3L7 AUST-02 subtypes and non AUST-02 strains.

Clinical Outcomes
The mean (±SD) follow-up period for the entire cohort was 1067 (±242) days. During this time, 6/166 (3.6%) patients died and 17/166 (10.2%) underwent lung transplantation (Fig 3). This 2) adjusted for the effects of age, BMI and severity of airway disease (Fig 4). In contrast, no differences were observed in the 3-year adjusted HR of death/lung transplantation between patients infected with non-M3L7 AUST-02 subtypes and non-AUST-02 strains. While none died within the M3L7 group, two died from respiratory failure in the non-M3L7 AUST-02 subtype group (one was ineligible for transplantation because of B. cepacia complex, the other declined a transplant) and another four (4%) died in the non AUST-02 group. Three of the four deaths in the non AUST-02 group were attributable to severe CFrelated airways disease, including two who died on the transplant waiting list and another who was ineligible for transplantation due to co-infection with B. cepacia complex. The other death in the non AUST-02 group was from metastatic gastrointestinal malignancy.
The secondary outcomes during the 3-year follow-up for the three patient groups are presented in Table 4. Adults harbouring M3L7 in 2007-2009 had greater rates of decline in BMI and lung function, although these did not reach statistical significance. It would also appear from the univariate analysis that those infected with M3L7 had a significantly greater number of hospitalisation days for IV antibiotics annually than either of the other two patient groups (Table 4).

Discussion
Microevolution of the P. aeruginosa genome is an important pathoadaptive process within the CF airways, characterised by extensive genotypic and phenotypic diversification [16,17,19,20]. However, the potential clinical implications of such ongoing microevolution within common CF-specific P. aeruginosa shared strains in terms of cross-transmission and impact on    clinical outcomes remains unclear. Consistent with earlier studies, we observed mexZ and LasR sequence diversity among the entire 188 isolate collection [16,21,22,29,30]. In particular, we observed the increased frequency of AUST-02 LasR sequence variants in adult compared with paediatric patients. Likewise, longitudinal changes in LasR sequence were also observed in isolates from approximately 40% of study patients. This observation is supported indirectly by other studies linking lasR loss-of-function mutations with the transition from early to chronic P. aeruginosa infection and accelerated pulmonary function decline in CF [22,25,30]. The lasR gene is a key transcriptional regulator of quorum sensing in P. aeruginosa and its mutational inactivation is associated with attenuated virulence, host immune evasion and a growth advantage from utilising the specific carbon and nitrogen sources available in the amino-acid rich CF airway secretions [31]. Inactivation of the lasR gene is also associated with increased βlactamase activity and ceftazidime tolerance [31], which facilitates its adaptation to the CF airway.
During the 10-year study period we observed the emergence and expansion of an intrastrain AUST-02 subtype (M3L7), characterised by an enhanced MDR phenotype, which superseded its ancestral wild-type AUST-02 strain and other AUST-02 subtypes. M3L7 appears to have retained its transmissibility, which presumably contributed to the additional incident cases observed within the adult CF clinic between 2007 and 2011. Our results also demonstrated significant differences in clinical trajectories between patients infected with different AUST-02 intra-strain subtypes.
As respiratory failure is the major cause of death in patients with CF, and lung transplantation is a viable treatment for end-stage CF lung disease, like others we chose time to either death or lung transplantation as the primary endpoint events in our survival analysis [11]. Infection with the M3L7 AUST-02 subtype was associated with a nine-fold increase in the overall 3-year risk of death or lung transplantation compared with other non-M3L7 AUST-02 subtypes and a five-fold increase in risk when compared with non AUST-02 strains. In contrast, no significant differences were detected between patients infected with non M3L7 AUST-02 subtypes and non AUST-02 strains. M3L7 infected patients also had significantly greater requirement for hospitalisation for IV antibiotics at baseline and follow-up. Therefore, our results suggested there might be differences in the clinical outcomes according to intra-strain subtypes within this commonly shared P. aeruginosa strain. The mechanisms underlying the association of the M3L7 subtype with poorer clinical outcomes remain uncertain. Possible explanations include its MDR phenotype could lead to less effective antibiotic treatment or it could simply represent a transmissible strain found more frequently in those with severe CF pulmonary disease following their increased healthcare requirements and exposure to other infected patients.
The latter is suggested by the lower baseline FEV 1 (% predicted) in the M3L7 infected patients compared with the other groups of P. aeruginosa infected patients, although with the small numbers involved this did not reach statistical significance. Whilst there were no deaths amongst M3L7 infected patients, a significantly greater proportion of M3L7 infected patients required lung transplantation than observed with the other groups. It is possible therefore that either the increased IV antibiotics requirement associated with M3L7 infection at both baseline and follow-up reflected the clinical impact of the MDR M3L7 subtype or, alternatively, the intensified IV antibiotic usage in the pre-transplant period or a combination of both. Furthermore, greater antibiotic exposure in patients with M3L7 infection may have also contributed to the development of the MDR phenotype observed with this subtype. Whole genome sequencing studies are planned to explore these possibilities.
Nevertheless, until the above questions are resolved, the emergence and ongoing transmission of this AUST-02 subtype raises important questions over current infection control practices. There is now evidence that P. aeruginosa survives in cough aerosols up to 4-metres and for as long as 45-minutes, suggesting cough bioaerosols as a possible transmission pathway [32][33][34]. Strain-specific cohort segregation practiced by some Australian and international CF centres has interrupted acquisition of their targeted strains [35][36][37]. However, as reported previously with B. cenocepacia [38], cohort-segregation without additional strain-typing risks acquiring more virulent (epidemic) subtypes. Our results agree with recent reports of crossinfection with intra-strain sublineages of the highly successful LES strain in CF patients cohorted according to strain-based segregation policies at an English CF centre [39]. We provide further observations of possible adverse clinical outcomes associated with the spread of an emerging MDR intra-strain subtype of AUST-02 at an Australian CF centre. These findings support the United Kingdom CF Standards of Care [40] and CF Foundation Infection Control Guidelines [41] that recommend all hospitalised patients remain in single rooms to minimise cross-infection. Australian CF centres are now optimising hospital-based and outpatient care to limit contact between patients.
This study does however have some important limitations. First, the P. aeruginosa isolates originated from three different projects. Whilst none of these studies were designed initially to monitor microevolution of bacterial strains, combining these historical sample banks has provided us with an ideal opportunity to examine the pattern of intra-strain diversification within the AUST-02 strain using a large number of isolates collected from a representative population of Queensland paediatric and adult CF patients infected over a decade. Second, as a maximum of three colonies were selected from sputum culture plates, the numbers of strains and their subtypes within individuals may have been substantially underestimated. However, we believe the limitation in the sampling depth is partially compensated by the extent and period of sampling, which has enabled us to assess the between-patient, rather than within-patient, AUST-02 population heterogeneity. Furthermore, a small number of patients appeared to have undergone repeated new strain acquisition at the three different time-points (2001-2002; 2007-2009; 2011) (Table 3). Given most patients retained the same strains for prolonged periods, this observation may relate to either repeated strain replacements from multiple cross-infection events or under-sampling of co-infecting strains. Thirdly, the relatively small number of patients infected with the M3L7 subtype is reflected in the broad 95%CIs around risk estimates, which means results must be interpreted cautiously. In addition, due to the paucity of P. aeruginosa strain data prior to the ACPinCF study (2007)(2008)(2009), the date of individual M3L7 acquisition remains unknown. Thus our outcome analysis was limited to prevalent rather than incident cases. Further prospective longitudinal studies are now needed to help clarify whether acquiring the M3L7 strain is associated with adverse clinical outcomes. Finally, we used mexZ and lasR sequencing as a fine typing tool to detect intra-strain diversity. While these two genes are frequent mutation targets contributing to early adaptive evolution [16,21,22,29] we observed limited sequence diversity in the mexZ gene in AUST-02 strains. Such restricted genomic diversity agrees with reports of another persistent, shared CF strain (DK2) from Denmark, which suggest that mutations in P. aeruginosa strains highly adapted to the CF lung are constrained [22,42]. Further study of these successful CF strains requires whole genome sequencing techniques that provide greater resolution than simply targeting selected genes.
In summary, our study showed the emergence and suspected person-to-person transmission of a MDR intra-strain subtype (M3L7) within AUST-02, the dominant P. aeruginosa strain in our adult CF clinic. This subtype was associated with poorer clinical outcomes than its ancestral strain. These data also suggest that commonly shared P. aeruginosa strains can evolve into more resistant subtypes, while also retaining the ability to infect other CF patients. This has important infection control implications when adopting P. aeruginosa strain-specific segregation strategies in the CF clinic. Based on the results from our current study and other studies from our group [13], a new purpose-designed CF ward with single room facilities for all CF patients hospitalised for IV antibiotics was commissioned in 2014, allowing total segregation of all patients from one another during hospitalisation. Studies are planned to characterise the underlying mechanisms associated with resistance, virulence and the metabolic profile of this newly described AUST-02 subtype.