Phenotypic diversity within a Pseudomonas aeruginosa population infecting an adult with cystic fibrosis

Abstract

Chronic airway infections caused by Pseudomonas aeruginosa contribute to the progression of pulmonary disease in individuals with cystic fibrosis (CF). In the setting of CF, within-patient adaptation of a P. aeruginosa strain generates phenotypic diversity that can complicate microbiological analysis of patient samples. We investigated within- and between- sample diversity of 34 phenotypes among 235 P. aeruginosa isolates cultured from sputum samples collected from a single CF patient over the span of one year, and assessed colony morphology as a screening tool for predicting phenotypes, including antimicrobial susceptibilities. We identified 15 distinct colony morphotypes that varied significantly in abundance both within and between sputum samples. Substantial within sample phenotypic heterogeneity was also noted in other phenotypes, with morphotypes being unreliable predictors of antimicrobial susceptibility and other phenotypes. Emergence of isolates with reduced susceptibility to β-lactams was observed during periods of clinical therapy with aztreonam. Our findings confirm that the P. aeruginosa population in chronic CF lung infections is highly dynamic, and that intra-sample phenotypic diversity is underestimated if only one or few colonies are analyzed per sample.

Figure 1. Temporal diversity of P. aeruginosa colony morphologies found in the sputum of a CF patient.

Diversity was assessed through (A) culture-dependent identification and (B) characterization of colony morphologies by morphotype over 350 days.

Figure 2. Diversity of adaptive phenotypes.

Fluctuations in motility-associated phenotypes (A-D) are represented by individually colour-coded morphotype, while those observed for amino acid auxotrophy (E), pyocyanin production (F), iridescent surface sheen (G), protease production (H), and siderophore production (I), are depicted at the population level. Phenotypes were measured among twenty isolates from each sputum sample with the exception of days 0 (19 isolates), 232 (18 isolates) and 350 (18 isolates). Each data point in panels A-C represents the mean of three replicates and six replicates in panel D, with standard deviation removed for presentation.

Figure 3. Variation in antimicrobial susceptibility patterns over time.

Changes in (A) antibiotic exposure and in vitro responses to (B) aztreonam, (C) ceftazidime, and (D) ciprofloxacin. Susceptibilities are expressed as the percentage of isolates recovered from a given sample with a particular response phenotype and defined as susceptible (MIC ≤ 8 μg/mL for ATM, CAZ; ≤ 1 μg/mL for CIP), intermediate (MIC of 16 μg/mL for ATM, CAZ; 2 μg/mL for CIP) or resistant (MIC ≥ 32 μg/mL for ATM, CAZ; ≥ 4 μg/mL for CIP) by CLSI interpretive criteria. Delivery routes for antimicrobial therapies are defined as inhaled (INH), parenteral (IV) or oral (PO).

Figure 4. Variation in β-lactam susceptibility within and between morphotypes.

The frequency of a particular antimicrobial susceptibility profile (for the antibiotics aztreonam and ceftazidime) for any morphotype is depicted by the size of the corresponding circle, representing low (single isolate with the susceptibility profile), moderate (2–3 isolates with the susceptibility profile) or high ( ≥ 4 isolates with the susceptibility profile) frequencies.