Article pubs.acs.org/jpr
Secretome of Transmissible Pseudomonas aeruginosa AES-1R Grown in a Cystic Fibrosis Lung-Like Environment Nichollas E. Scott,†,∥,⊥ Nathan J. Hare,†,∥ Melanie Y. White,† Jim Manos,‡ and Stuart J. Cordwell*,†,§ †
School of Molecular Bioscience, The University of Sydney, Building GO8, Maze Crescent, Sydney, New South Wales 2006, Australia Discipline of Immunology and Infectious Diseases, The University of Sydney, Room 667, Level 6 Blackburn Building D06, Sydney, New South Wales 2006, Australia § Discipline of Pathology, School of Medical Sciences, The University of Sydney, Room 501, Level 5 Blackburn Building D06, Sydney, New South Wales 2006, Australia ‡
S Supporting Information *
ABSTRACT: Pseudomonas aeruginosa is the predominant cause of mortality in patients with cystic fibrosis (CF). We examined the secretome of an acute, transmissible CF P. aeruginosa (Australian epidemic strain 1-R; AES-1R) compared with laboratory-adapted PAO1. Culture supernatant proteins from rich (LB) and minimal (M9) media were compared using 2-DE and 2DLC-MS/MS, which revealed elevated abundance of PasP protease and absence of AprA protease in AES-1R. CF lung-like artificial sputum medium (ASMDM) contains serum and mucin that generally preclude proteomics of secreted proteins. ASMDM culture supernatants were subjected to 2DLC-MS/MS, which allowed the identification of 57 P. aeruginosa proteins, and qualitative spectral counting was used to estimate relative abundance. AES-1R-specific AES_7139 and PasP were more abundant in AES-1R ASMDM culture supernatants, while AprA could only be identified in PAO1. Relative quantitation was performed using selected reaction monitoring. Significantly elevated levels of PasP, LasB, chitin-binding protein (CbpD), and PA4495 were identified in AES-1R ASMDM supernatants. Quantitative PCR showed elevated pasP in AES-1R during early (18 h) ASMDM growth, while no evidence of aprA expression could be observed. Genomic screening of CF isolates revealed aes_7139 was present in all AES-1 and one pair of sequential nonepidemic isolates. Secreted proteins may be crucial in aiding CF-associated P. aeruginosa to establish infection and for adaptation to the CF lung. KEYWORDS: artificial sputum medium, cystic fibrosis infection, extracellular proteins, secretome, selected reaction monitoring (SRM)
■
INTRODUCTION Pseudomonas aeruginosa is a Gram-negative opportunistic human pathogen with a comparatively large genome (∼5500 genes1) that confers an extraordinary ability to adapt to a variety of environments. Nosocomial P. aeruginosa infections include those involving burns and wounds, although meningitis, endocarditis, and microbial keratitis are also possible. P. aeruginosa is the major determinant of morbidity and mortality in patients suffering from the autosomal recessive disorder cystic fibrosis (CF).2 CF affects ∼1 in 3300 live births in the Caucasian population.3 The genetic lesion is located within the CF transmembrane conductance regulator (CFTR), a chloride channel expressed in epithelial cells. While over 1000 cf tr mutations have been reported, the ΔF508 form is predominant.4,5 Defects in CFTR lead to electrolyte transport imbalance across many organ systems. In the lungs, this is characterized by the production of a thickened, dehydrated mucus layer that provides an environment suitable for colonization by opportunistic pathogens.2 Although many species are able to colonize the CF lung, P. aeruginosa will eventually dominate in the majority of patients. Initial infection by nonmucoid P. aeruginosa during early childhood can be intermittent and cleared by antibiotics; © 2013 American Chemical Society
however, by adulthood more than 80% of patients harbor mucoid strains that cannot be eradicated by even aggressive antibiotic therapy.6 Conversion to mucoidy (overproduction of alginate capsular polysaccharide) confers a selective advantage and is an indicator of poor prognosis.7 Chronic P. aeruginosa infection and persistent inflammation lead to a progressive decline in lung function and eventual death. It was previously thought that CF patients acquire P. aeruginosa from the environment, and person−person transmission, even among siblings, was considered rare. The emergence of transmissible “epidemic” CF strains has, however, now been observed in clinics across the U.K., Australia (Australian epidemic strains 1, 2, and 3 [AES-1,-2,-3]), Brazil, and Denmark.8−16 Epidemiological studies strongly suggest that transmission does not occur through shared contact with an environmental source but via direct aerosol between patients.17,18 Transmissible strains are also associated with worse prognosis, suggesting enhanced virulence properties.19 Received: September 24, 2012 Published: August 30, 2013 5357
dx.doi.org/10.1021/pr4007365 | J. Proteome Res. 2013, 12, 5357−5369
Journal of Proteome Research
Article
CF-associated P. aeruginosa grow initially as free cells during early infection and then as microcolonies resembling a biofilm phenotype where cells aggregate but do not attach to surfaces; such biofilms confer additional resistance to chemicals and antibiotics. In vivo CF models preclude proteomics analysis, and hence in vitro culture models are necessary. Such models include solid innate surfaces for biofilm simulation, either submerged within medium (e.g., glass wool model20), with circulating medium (e.g., continuous-flow biofilm reactors21), or in “sputum”-like medium (e.g., agar-entrapped cultures22 or artificial sputum medium (ASM)23,24). The addition of CF patient sputum is impractical because it is difficult to acquire and sterilize, may contain contaminating yeasts, varies greatly between patients, and may reflect individual antibiotic treatment. ASM is therefore an excellent model because it simulates the viscosity of CF sputum, contains a gradient of oxygen availability, and allows microcolony formation24 without the need for patient sputum, although the presence of serum proteins, DNA, and mucin may complicate proteomic analysis. P. aeruginosa encodes secreted proteins,25−28 including proteases and toxins as well as small molecule siderophores, such as pyochelin and pyoverdine29,30 that sequester low-level nutrients, particularly iron, from the environment. Other secreted secondary metabolites, most notably pyocyanin, are also major virulence determinants. Secreted virulence factors under the control of population-dependent quorum sensing (QS) include elastase LasB, LasA protease, alkaline protease AprA, chitin-binding protein (CbpD), chitinase, PasP protease, and PrpL endoproteinase.26,28 The secretome is therefore likely to be a major determinant in P. aeruginosa transmission to, and colonization of, the CF lung. Previous studies have examined secreted proteins from P. aeruginosa strains associated with CF (e.g., strain C31); however, extracellular profiles have typically been generated by growth in standard laboratory media that do not resemble CF lung conditions. Examination of secreted proteins from non-CF P. aeruginosa PAO1 using sera from chronically infected CF patients has, however revealed the identities of antigens, including LasB, azurin, PrpL, PasP, and PA2939.32 In this study, secreted proteins from stationary phase P. aeruginosa PAO1 and the acute, transmissible CF isolate, AES1R,33,34 were examined in culture supernatants from nutrientrich (Luria−Bertani [LB]) and nutrient-limited (M9) media and proteins examined by 2-DE and 2DLC-MS/MS. Both strains were then grown in modified ASM (ASM with low-molecularweight DNA and mucin; ASMDM24,35) and proteins examined by 2DLC-MS/MS. We identified 57 proteins across the two P. aeruginosa strains despite interference from serum proteins and mucin. The relative abundance of secreted proteins in ASMDM was compared between PAO1 and AES-1R by selected reaction monitoring (SRM) assays for 10 proteins. P. aeruginosa AES-1R produces high levels of secreted virulence factors and hypothetical proteins that may be important in establishing infection in the CF lung.
■
strains LES431, LESB58, Man C3733, Man 8799, Mid 10066, Mid 8916, Clone C, Stoke, Trent, Leiden, NE43, and NE127 were kindly provided by Prof. Craig Winstanley, Institute of Infection and Global Health, University of Liverpool. P. aeruginosa PAO1 and AES-1R were cultured in six replicates each of (i) 50 mL of nutrient-rich salt-modified LB broth (5 g/L NaCl) and (ii) 50 mL of nutrient-limited M9 minimal medium supplemented with 2% (w/v) glucose to stationary phase (OD600 nm ≈ 1.0) with incubation at 37 °C and shaking at 250 × rpm. Cultures were harvested and washed three times with PBS. ASMDM cultures were carried out as per ref 24 with minor modifications as described in ref 35, including supplementation with 3 μg/mL horse spleen ferritin (Sigma, St. Louis, MO) to reflect carrier protein-bound iron levels in CF sputum (normal control 0.2 mg/L, acute CF 3.6 mg/L, and stable CF 2.4 mg/ L36). Fifteen ASMDM cultures for each strain were carried out in parallel, and the experiment was replicated three times. Cultures were incubated statically at 37 °C with a loosened lid for 72 h. Preparation of P. aeruginosa Secreted Proteins and 2-DE
For P. aeruginosa strains grown in LB and M9, cells were removed by centrifugation at 6000g for 10 min at 4 °C, and the resulting supernatants were filtered using 0.22 μm syringe filters (Millipore, Bedford, MA). The filtrates were transferred to snakeskin 3500 Da molecular weight dialysis membranes (Pierce, Rockford, IL) and dialyzed against 5 L of ultrapure water for 3 h at 4 °C. The procedure was then repeated twice with fresh ultrapure water. Samples were transferred to fresh tubes in 35 mL aliquots and freeze-dried overnight. Dried proteins from LB and M9 media were resuspended in 1 mL of 2-DE buffer (5 M urea, 2 M thiourea, 0.1% (v/v) carrier ampholytes, 2% (w/v) CHAPS, 2% (w/v) sulfobetaine 3−10, 2 mM tributylphosphine). 250 μg of protein was used to reswell precast 17 cm pH 4−7 immobilized pH gradient (IPG) strip gels (Bio-Rad). The same protein load was added via anode cup-loading to pH 6−11 IPG strips37 (GE Healthcare, Uppsala, Sweden). Isoelectric focusing, reduction/alkylation and detergent exchange, and second-dimension SDS-PAGE were performed as described.34,35 Gels were fixed in 40% (v/v) methanol and 10% (v/v) acetic acid for 1 h and then stained overnight in Sypro Ruby (Bio-Rad). Gels were destained in 10% (v/v) methanol and 7% (v/v) acetic acid for 1 h and imaged using a Molecular Imager Fx (Bio-Rad). Gels were “double-stained” in Colloidal Coomassie Blue G-250. Image and statistical analysis of protein spot densities were performed using PD-Quest (Bio-Rad), as previously described34,35 (see Supplementary Methods in the Supporting Information). Identification of Proteins from 2-DE Gels Using MALDI-TOF MS
Protein spots were digested with trypsin, concentrated and desalted, and analyzed by MALDI-TOF MS as previously described.35 Data from peptide mass maps were used to perform searches of a composite P. aeruginosa database composed of translated genome sequences from PAO1 (Pseudomonas Genome Database v2, 2009-11-23; 18 694 entries), PA14 (Pseudomonas Genome Database v2, 2009-10-14), and AES1R33 via an in-house MASCOT server (Matrix Science; version 2.2), as described in refs 34 and 35 (see Supplementary Methods in the Supporting Information).
MATERIALS AND METHODS
Bacterial Strains and Growth Conditions
P. aeruginosa strains used in this study are listed in Suppl. Table S1 in the Supporting Information. P. aeruginosa PAO1 (ATCC 15692) was obtained from the American Type Culture Collection. P. aeruginosa AES-1R was obtained from David Armstrong, Monash Medical Centre, Australia.8 P. aeruginosa
2DLC-MS/MS of Secreted Proteins from P. aeruginosa
Dried secreted proteins (500 μg) from P. aeruginosa PAO1 and AES-1R grown in LB, M9, or ASMDM were resuspended in 6 M 5358
dx.doi.org/10.1021/pr4007365 | J. Proteome Res. 2013, 12, 5357−5369
Journal of Proteome Research
Article
autosampler system (LC Packings, Amsterdam, Netherlands). Samples (5 μL) were concentrated and desalted on a micro C18 precolumn (500 μm × 2 mm, Michrom Bioresources, Auburn CA) with Buffer A (2% MeCN, 0.1% TFA) at 15 μL/min. After a 4 min wash, the precolumn was switched (Switchos, LC Packings) into line with a fritless C18 nano column (5 μ 100 Å Magic, 75 μm × ∼10 cm, Michrom). Peptides were eluted using a linear gradient of Buffer A (2% MeCN, 0.1% formic acid) to (62% MeCN, 0.1% FA) at ∼300 nL/min over 90 min. High voltage (2400 V) was applied, and the column tip was positioned ∼1 cm from the orifice (T = 150 °C) of a 5500 Q-TRAP hybrid triple quadrupole/linear ion trap mass spectrometer (AB Sciex) operating with Analyst v2.0. Transition lists generated by Skyline were acquired with a 10 ms dwell time, and five unscheduled replicates of each sample were analyzed. The resulting data files were imported into Skyline, where the transition boundaries were inspected and areas under the curve were exported for statistical analysis using Student’s t test. All transitions, representative extracted ion chromatograms, and statistical data are provided in Supporting Information (Suppl. Data S1). To ensure that transitions were indicative of the peptide-of-interest, we subjected peptides to MS/MS (Suppl. Data S1 in the Supporting Information), and MASCOT searches were performed against the Ludwig nonredundant (to ensure the selected peptide was not from a contaminating media-related protein) and P. aeruginosa databases. Only peptides that resulted in accurate identification (mass tolerance of 1 Da and no missed cleavages) were retained for quantitation.
urea, 2 M thiourea, and 40 mM NH4HCO3 and analyzed by 2DLC-MS/MS as previously described34,35 (see Supplementary Methods in the Supporting Information). Peptides were analyzed using an LTQ-Orbitrap XL mass spectrometer (Thermo Scientific, San Jose, CA). The instrument was operated in a data-dependent mode automatically selecting the top five most intense ions with a charge state of +2 or greater for MS/MS. The selected ions were subjected to CID fragmentation with normalized collision energy of 35 and with dynamic exclusion enabled and set to 45 s. MS/MS spectra were extracted by Proteome Discoverer vers. 1.0 Build 43 (Thermo Scientific) and converted to Mascot generic format (.mgf) files. All MS/MS spectra were analyzed using MASCOT and X!Tandem (www. thegpm.org; version 2007.01.01.2). Searches were performed using the in-house MASCOT server against the P. aeruginosa database described above. Search parameters included one possible missed cleavage and the following variable modifications: deamidation of Asn/Gln and oxidation of Met and carbamidomethyl-Cys. MASCOT and X!Tandem were searched with a fragment ion mass tolerance of 0.8 Da and a parent ion tolerance of 10 ppm. Scaffold (v. Scaffold_3_00_03, Proteome Software, Portland, OR) was used to validate MS/MS-based peptide and protein identifications. Peptide identifications were accepted if they could be established at >95% probability, as specified by the Peptide Prophet algorithm.38 MASCOT identifications required ion scores greater than both the associated identity scores and with values of at least 30, 40, and 50 for doubly, triply, and quadruply charged peptides, respectively. X!Tandem identifications required −Log (expect scores) scores of 99.0% probability and contained at least one identified peptide. Protein probabilities were assigned by Protein Prophet.39 Proteins that contained similar peptides that could not be differentiated based on MS/MS analysis alone were grouped to satisfy the principles of parsimony. The protein false discovery rate was determined by Scaffold to be 0.1%. MS/MS spectra assigned to a P. aeruginosa protein were manually annotated according to ref 40 to further ensure data integrity and considering the relatively low numbers of peptides
Quantitative Real-Time Reverse Transcriptase PCR (qRT-PCR)
P. aeruginosa AES-1R and PAO1 were grown in ASMDM to 18 and 72 h, which are reflective of early (acute) and microcolony (chronic) stages of CF lung infection. RNA was extracted using an RNeasy Mini Kit (Qiagen, Valencia, CA) with RNAprotect to prevent degradation and RNase-free DNase treatment to remove genomic DNA. RNA concentration was determined by absorbance at 260 nm, and quality was assessed by absorbance ratio A260/A280 and formaldehyde gel electrophoresis. A total of 100 ng of RNA was used for cDNA synthesis using the Superscript VILO kit (Invitrogen, Carlsbad CA). Real-time RTPCR primers for pasP and aprA were synthesized (SigmaAldrich), and qRT-PCR was performed as previously described35 (Suppl. Table S2 in the Supporting Information). Agarose gel electrophoresis was also carried out on PCR products to ensure single products. Total cDNA abundances between samples were normalized using 16S rRNA as a control. qRT-PCR was carried out in duplicate on three biological replicates. Student’s t test was performed to determine statistical significance.
Selected Reaction Monitoring of P. aeruginosa Secreted Proteins in ASMDM
Proteotypic peptides for use in SRM were chosen based on both observed data generated by 2DLC-MS/MS and in silico determination. Peptides were chosen only if tryptic and were excluded if they contained a missed cleavage or variable modification, such as Met oxidation. Preference was given to peptides containing amino acids known to produce intense fragment ions (e.g., Pro and Asp). Identified peptides and their six most intense fragment ions were evaluated. Where the available identified peptides did not meet these criteria, in silico peptide/transition selection was accomplished using Skyline v0.6.41 Transition combinations were optimized to ensure that multiple ions were detected, and a final CV of
