A Shaving Proteomic Approach to Unveil Surface Proteins Modulation of Multi-Drug Resistant Pseudomonas aeruginosa Strains Isolated From Cystic Fibrosis Patients

Abstract

Cystic fibrosis (CF) is the most common rare disease caused by a mutation of the CF transmembrane conductance regulator gene encoding a channel protein of the apical membrane of epithelial cells leading to alteration of Na⁺ and K⁺ transport, hence inducing accumulation of dense and sticky mucus and promoting recurrent airway infections. The most detected bacterium in CF patients is Pseudomonas aeruginosa (PA) which causes chronic colonization, requiring stringent antibiotic therapies that, in turn induces multi-drug resistance. Despite eradication attempts at the first infection, the bacterium is able to utilize several adaptation mechanisms to survive in hostile environments such as the CF lung. Its adaptive machinery includes modulation of surface molecules such as efflux pumps, flagellum, pili and other virulence factors. In the present study we compared surface protein expression of PA multi- and pan-drug resistant strains to wild-type antibiotic-sensitive strains, isolated from the airways of CF patients with chronic colonization and recent infection, respectively. After shaving with trypsin, microbial peptides were analyzed by tandem-mass spectrometry on a high-resolution platform that allowed the identification of 174 differentially modulated proteins localized in the region from extracellular space to cytoplasmic membrane. Biofilm assay was performed to characterize all 26 PA strains in term of biofilm production. Among the differentially expressed proteins, 17 were associated to the virulome (e.g., Tse2, Tse5, Tsi1, PilF, FliY, B-type flagellin, FliM, PyoS5), six to the resistome (e.g., OprJ, LptD) and five to the biofilm reservoir (e.g., AlgF, PlsD). The biofilm assay characterized chronic antibiotic-resistant isolates as weaker biofilm producers than wild-type strains. Our results suggest the loss of PA early virulence factors (e.g., pili and flagella) and later expression of virulence traits (e.g., secretion systems proteins) as an indicator of PA adaptation and persistence in the CF lung environment. To our knowledge, this is the first study that, applying a shaving proteomic approach, describes adaptation processes of a large collection of PA clinical strains isolated from CF patients in early and chronic infection phases.

Keywords: Pseudomonas aeruginosa; antibiotic resistance; cystic fibrosis; long-term colonization; shaving proteomics.

Figure 1

Graph distribution of purified peptides and identified proteins. (A) After the shaving procedure, a mean of 6.27 μg (±2.60 standard deviation), 5.72 μg (±1.60) and 5.21 μg (±1.40) of peptides in the MDR, PDR, and WT samples, respectively, were purified. (B) By nLC–ESI–MS/MS, we identified a mean of 801 (±379), 1,117 (±221), and 950 (±208) proteins were identified in the MDR, PDR, and WT groups, respectively.

Figure 2

Analyses of quantified proteins. (A) The dissimilarity between samples' groups by the unsupervised Bray-Curtis beta diversity analysis was measured according to their proteins' content. MDR, PDR and WT groups showed a statistically significant (p-value ≤ 0.001) differences assessed by PERMANOVA test (PC1 vs. PC2; PC1: 31.47%; PC2: 20.75%). (B) Unsupervised Principal Component Analysis displayed a good separation amongst groups (PC1 vs. PC2; PC1: 19.93%; PC2: 15.98%). (C) Color-coded hierarchical cluster analysis visualized by the heat map based on the normalized protein abundances and applying a z-score transformation. The dendrogram above the heat map, representing the distance between samples, demonstrated good similarity among samples of each strain groups. The top left heat map color legend displays the range of the scaled protein abundance values, ranging from −2.1 to +4.9 and a mean of +1.4. Blue, orange and green color of (A–C) labels/resistance profiles correspond to MDR, PDR and WT P. aeruginosa strains, respectively.

Figure 3

Number of quantified proteins and their subcellular localization, expressed as percentage respect to the total number of quantified proteins, by merging data from several web-based applications (PSORTb-3.0, SignalP-5.0, TargetP-2.0, LipoP 1.0 CELLO v.2.5, SecretomeP-2.0, TOPCONS, TMHMM-2.0, DeepTMHMM, UniProtKB database and Proteome Discoverer ProteinCenter annotation).

Figure 4

Venn diagram showing common and specific non-cytosolic (extracellular space, cell surface, outer membrane, periplasmic region, and cytoplasmic membrane) proteins in MDR vs. WT (119 total proteins) and PDR vs. WT (108 total proteins) comparisons based on label-free quantification. Percentages of the distribution among the different subcellular regions, respect the total quantified proteins, are reported below each subset of the diagram.

Figure 5

Sketch of two secretion systems (Type I and VI), transporting microbial substrates across membranes of Gram-negative bacteria in one step. Image modified from pae03070 KEGG (release 99.0, July 1, 2021) pathway depicting participating factors, proteins, complexes, and their subcellular distribution. MDR and PDR P. aeruginosa clinical strain identified proteins are evidenced by dashed box along with their shift of expression compared to WT strains (red dashed box and downward arrow = under-expression; green box and upward arrow = over-expression). Probable outer membrane protein (UniProtKB code Q9HUJ1, KEGG identifier PA4974) is localized in the outer membrane and involved in bacterial secretion system Type I. The identified Uncharacterized protein (Q9I1B2) is a homologous of Hcp1 family type VI secretion system effector and is mapped by KEGG software (KEGG identifier PA2367) as type VI secretion system secreted protein Hcp ortholog. Toxin Tse2 (Q9I0E0), Toxin protein Tse5 (Q9I0F4) and Immune protein Tsi1 (Q9I2Q0) are other members of the type VI secretion system.

Figure 6

Sketch of the β-Lactam resistance pathway. Image modified from pae01501 KEGG (release 99.0) pathway showing involved factors/effectors, their subcellular distribution and their molecular interactions, reactions and relation networks. MDR and PDR P. aeruginosa clinical strain identified proteins are represented by a dashed box along with their shift of expression compared to WT strains (red dashed box and downward arrow = under-expression). Among the outer membrane proteins, we identified a Probable outer membrane protein (UniProtKB code Q9HUJ1, KEGG identifier PA4974) and the Outer membrane protein OprM (Q51487, PA0427), both involved in the resistance-nodulation division (RND) efflux pump mechanisms at their outer membrane subcellular localization. OprM modulation may also result from intrinsic properties of organisms, through gene mutations, and through plasmid- and transposon-specified genes at the DNA level (OpMAdeK system). Light blue box = input protein, differentially expressed in MDR or PDR vs. WT. Green box = P. aeruginosa-specific pathway. Red characters = resistance associated gene variants.