Genomic and Functional Characterization of Longitudinal Pseudomonas aeruginosa Isolates from Young Patients with Cystic Fibrosis

Abstract

Individuals with cystic fibrosis (CF) suffer from frequent and recurring microbial airway infections. The Gram-negative bacterium Pseudomonas aeruginosa is one of the most common organisms isolated from CF patient airways. P. aeruginosa establishes chronic infections that persist throughout a patient's lifetime and is a major cause of morbidity and mortality. Throughout the course of infection, P. aeruginosa must evolve and adapt from an initial state of early, transient colonization to chronic colonization of the airways. Here, we examined isolates of P. aeruginosa from children under the age of 3 years old with CF to determine genetic adaptations the bacterium undergoes during this early stage of colonization and infection. These isolates were collected when early aggressive antimicrobial therapy was not the standard of care and therefore highlight strain evolution under limited antibiotic pressure. Examination of specific phenotypic adaptations, such as lipid A palmitoylation, antibiotic resistance, and loss of quorum sensing, did not reveal a clear genetic basis for such changes. Additionally, we demonstrate that the geography of patient origin, within the United States or among other countries, does not appear to significantly influence genetic adaptation. In summary, our results support the long-standing model that patients acquire individual isolates of P. aeruginosa that subsequently become hyperadapted to the patient-specific airway environment. This study provides a multipatient genomic analysis of isolates from young CF patients in the United States and contributes data regarding early colonization and adaptation to the growing body of research about P. aeruginosa evolution in the context of CF airway disease. IMPORTANCE Chronic lung infection with Pseudomonas aeruginosa is of major concern for patients with cystic fibrosis (CF). During infection, P. aeruginosa undergoes genomic and functional adaptation to the hyperinflammatory CF airway, resulting in worsening lung function and pulmonary decline. All studies that describe these adaptations use P. aeruginosa obtained from older children or adults during late chronic lung infection; however, children with CF can be infected with P. aeruginosa as early as 3 months of age. Therefore, it is unclear when these genomic and functional adaptations occur over the course of CF lung infection, as access to P. aeruginosa isolates in children during early infection is limited. Here, we present a unique cohort of CF patients who were identified as being infected with P. aeruginosa at an early age prior to aggressive antibiotic therapy. Furthermore, we performed genomic and functional characterization of these isolates to address whether chronic CF P. aeruginosa phenotypes are present during early infection.

Keywords: LPS evolution; Pseudomonas aeruginosa; airway adaptation; cystic fibrosis; genomics.

FIG 1

CF isolates from young children selected for sequencing. Isolates were selected from a total of nine different patients over time. In some cases, multiple isolates from the same time point were collected. OP, collection via oropharyngeal swab; BAL, collection via bronchial alveolar lavage; Hospitalization, samples collected during a hospital visit caused by an exacerbation event; Loc., location (region); Pat., patient. Region numbers: 1, Seattle, WA; 2, Houston, TX; 3, Cleveland, OH.

FIG 2

Phylogeny of newly sequenced P. aeruginosa isolates with available complete genomes of P. aeruginosa. All newly sequenced isolates are highlighted according to their sample origin (light blue, child CF isolate; dark blue, adult CF isolate; light purple, non-CF bronchiectasis; red, nonairway clinical disease isolates; yellow, PAO1; green, environmental). Downloaded whole-genome origins are indicated with circles along with their associated NCBI accession numbers.

FIG 3

Visualization of location/region-specific genes. Raw LS-BSR values for region-specific genes were visualized in a heat map. Specific gene annotations are listed in Table S3 in the supplemental material. Many of the region-specific genes that were identified appear to also be patient associated.

FIG 4

LS-BSR analysis of genes associated with antibiotic resistance and disc diffusion data. LS-BSR analysis of genes associated with antibiotic resistance (A) was coupled with disc diffusion data for several antibiotics (B): aztreonam (ATM 30), cefepime (FEP 30), cefoperazone (CFP 30), cefotaxime (CTX 30), imipenem (IPM 10), piperacillin-tazobactam (TPZ 110), carbenicillin (CAR 100), amikacin (AMK 30), gentamicin (CN 10), tetracycline (TE 30), chloramphenicol (C 30), colistin (CT 10), and ciprofloxacin (CIP 5). The numbers following the antibiotic abbreviations correspond to the antibiotic disc concentrations in micrograms. For reference, the patient age at the time of isolation and the isolation source are included. “*” indicates allelic variation.

FIG 5

LS-BSR analysis of QS-related genes. The top array shows results from LS-BSR analysis of quorum sensing (QS)-related genes from CF isolates of P. aeruginosa (top array). QS genes were identified in a literature search. The bottom array indicates PQS production, as previously assayed from passage 2 isolates. Fold change is respective to a P. aeruginosa mutant that does not express PQS.