Bacterial Detection Using Unlabeled Phage Amplification and Mass

Article pubs.acs.org/jpr

Bacterial Detection Using Unlabeled Phage Amplification and Mass Spectrometry through Structural and Nonstructural Phage Markers Armelle Martelet,†,‡ Guillaume L’Hostis,†,‡ Paulo Tavares,§ Sandrine Brasilès,§ François Fenaille,‡ Christine Rozand,† Alain Theretz,† Gaspard Gervasi,† Jean-Claude Tabet,∥ Eric Ezan,⊥ Christophe Junot,‡ Bruno H. Muller,*,†,‡,# and François Becher*,‡,# †

bioMérieux S.A., 376, Chemin de l’Orme, 69280 Marcy-l’Etoile, France CEA, iBiTec-S, SPI, Laboratoire d’Etude du Métabolisme des Médicaments (LEMM), Bâtiment 136, 91191 Gif-sur-Yvette, France § CNRS UPR3296 and IFR 115, Unité de Virologie Moléculaire et Structurale (VMS), Bâtiment 14B, CNRS, 91198 Gif-sur-Yvette, France ∥ Université Pierre et Marie Curie (Paris 6), UMR 7201, Equipe de Spectrométrie de Masse, Institut Parisien de Chimie Moléculaire, 4 place Jussieu, 75005 Paris, France ⊥ CEA, Service de Biochimie et Toxicologie Nucléaire (SBTN), BP 17171, 30207 Bagnols-sur-Cèze, France ‡

S Supporting Information *

ABSTRACT: According to the World Health Organization, food safety is an essential public health priority. In this context, we report a relevant proof of feasibility for the indirect specific detection of bacteria in food samples using unlabeled phage amplification coupled to ESI mass spectrometry analysis and illustrated with the model phage systems T4 and SPP1. Highresolving power mass spectrometry analysis (including bottom-up and top-down protein analysis) was used for the discovery of specific markers of phage infection. Structural components of the viral particle and nonstructural proteins encoded by the phage genome were identified. Then, targeted detection of these markers was performed on a triple quadrupole mass spectrometer operating in the selected reaction monitoring mode. E. coli at 1 × 105, 5 × 105, and 1 × 106 CFU/mL concentrations was successfully detected after only a 2 h infection time by monitoring phage T4 structural markers in Luria−Bertani broth, orange juice, and French bean stew (“cassoulet”) matrices. Reproducible detection of nonstructural markers was also demonstrated, particularly when a high titer of input phages was required to achieve successful amplification. This strategy provides a highly time-effective and sensitive assay for bacterial detection. KEYWORDS: unlabeled phage, SRM, bacterial detection, marker discovery, high-resolving power mass spectrometry, nonstructural protein

■

INTRODUCTION

methods are thus being continuously developed to enhance detection efficiency aiming to combine speed and simplicity of use.3 One with the widest application is the polymerase chain reaction (PCR).4 Although a powerful technology, a major drawback of PCR-based methods is that they do not discriminate between dead and living microorganisms, thus potentially increasing the number of false-positive results.5,6 Moreover, under harsh environmental conditions, nucleic acid amplification-based assays6 are prone to false-negative results due to the presence of enzymatic inhibitors in the sample. Therefore, alternative methods are being developed, such as bacteriophage-mediated detection assays.7

Foodborne pathogens are a growing concern regarding human illness and death. The World Health Organization estimates that each year up to one-third of the populations of developed countries are affected by foodborne illness and that the problem is likely to be significantly more widespread in developing countries.1 Therefore, microbial analysis of food is an integral part of the management of safety in the food chain. Pathogen detection technology is central for the prevention and identification of food problems impacting health and safety.2 Although culture-based methods are standard microbiological techniques to detect single targeted bacteria, growth amplification from cells to colonies takes days to yield conclusive results.2 This is an obvious drawback in many industrial applications, particularly in the food sector where fast-response approaches are often required. Alternative assay © 2014 American Chemical Society

Received: October 2, 2013 Published: February 11, 2014 1450

dx.doi.org/10.1021/pr400991t | J. Proteome Res. 2014, 13, 1450−1465

Journal of Proteome Research

Article

virus capsid (or head) protein. However, a limitation using phage amplification is the need to differentiate input phage particles from progeny particles in the assay. To overcome this limitation, Pierce et al.28 used metabolically labeled 15N input phages that can be easily differentiated from 14N-progeny phages. In addition, they coupled phage amplification to an electrospray ionization (ESI) MS/MS technique, called selected reaction monitoring (SRM), for the quantitation of S. aureus in a culture medium using the stable isotope dilution technique.29−31 First introduced in the late 1970s,32 SRM is an MS technique that relies on the ability of triple quadrupole mass spectrometers33 to act as mass filters to selectively isolate targeted analytes and one of its fragments. Targeted SRM34 analysis is emerging in the field of proteomics35 and offers high selectivity, sensitivity, and a wide dynamic range for quantitative analysis.36 However, it shall be noted that the workflow of an SRM-based method requires previous identification of the proteins of interest and of their proteotypic peptides, using, for instance, results from an initial proteomic study. Marker identification is a crucial step for the development of their subsequent targeted detection. Protein identification via MS is usually carried out either by analysis of whole-protein analysis (top-down proteomics) or enzymatically produced peptides (bottom-up proteomics).37 Here we describe an ESI-MS strategy to identify two protein families used as phage markers: structural proteins that are assembled in viral particles and nonstructural proteins that are encoded by the phage genome but absent of the infective particles. Then, we describe a new approach using unlabeled phage amplification for the fast and sensitive detection of bacteria in media and in complex food matrices in which phage markers are detected directly in supernatants of infected bacteria, followed by an SRM analysis. The well-known lytic phages T4 and SPP1 were chosen as models for the detection of Gram-negative (Escherichia coli) and Gram-positive (Bacillus subtilis) bacteria, respectively. Detection of E. coli bacteria in complex food matrices was performed down to 5 × 105 CFU/ mL for the first time through the SRM monitoring of phage T4specific structural proteins. The usefulness of monitoring nonstructural markers of phage SPP1 is also apparent when a high titer of input phages is required to achieve successful amplification. The interest of each type of markers for bacterial detection using unlabeled phage amplification is discussed.

Bacterial viruses (bacteriophages or phages) are obligate parasites of bacteria using the resources of the bacterial cell to replicate.8 Successful phage multiplication requires metabolically active bacteria allowing differentiation between live and dead host cells. To survive and to propagate for millions of years, phages adapted and evolved to withstand harsh environmental conditions combined with the capacity to infect their bacterial hosts in various conditions9 (culture, water, clinical, and environmental matrices).7,10,11 Some phages possess a broad host range12 multiplying in numerous bacteria across strains, species, or genera. Other phages possess a very narrow host range and may infect only a limited number of bacteria within a particular strain.10 This feature is of particular interest for detection in samples where low numbers of pathogenic bacterial species are present together with many other nonpathogenic organisms. Phage specificity was demonstrated, for example, by the commercial VIDAS UP tests (bioMérieux, Marcy l’Etoile, France) for the detection of Salmonella and E. coli O157 (including H7),13 which use phage proteins as capture ligands. Altogether, it has become widely recognized in recent years that phages have a high potential for detecting pathogenic bacteria throughout the food chain.14 Phage-mediated detection approaches are very diverse such as phage typing, reporter phage, and phage amplification assays.7 More specifically, phage amplification assays exploit the natural infection cycle of phages and rely on measurement of progeny phages released from target host cells following cell lysis. An increase in the number of phages can therefore be used as an indicator of a successful infection event.10 Phages are used not only for their specificity to detect unwanted viable pathogens in foods but also to allow signal amplification that results from the increase in phage particle concentration and in phage proteins synthesized to a high copy number during infection. A single bacterial cell can produce hundreds of phage particles that are assembled from multiple copies of different structural proteins (up to 1000 copies per particle for some components). Phage multiplication in host cells is a complex process that also engages numerous nonstructural proteins, which are encoded by the phage genome but nonassembled in viral particles (as opposed to structural proteins).15 The phage amplification method was first described by Hirsh and Martin in 1983 for the detection of Salmonella spp. in pure culture medium and in milk.16 After Salmonella-specific phage Felix-01 amplification, the progeny phage particles were detected by high-performance liquid chromatography (HPLC) coupled to a UV measurement. Nowadays, phage amplification assays have been extended to other applications, and there are many different techniques for the detection of progeny phages such as plaque assays for the rapid diagnosis of tuberculosis,17 immunoassays for the screening of methicillin-resistant Staphylococcus aureus (MRSA),18 and quantitative real-time PCR for the detection of Yersinia pestis.19 Mass spectrometry (MS) was also used as a detection method for phage amplification assays.20−22 MS has become the method of choice for the analysis of complex protein samples,23,24 thus constituting a particularly useful tool for the identification and characterization of microorganisms.25−27 It offers the advantage of a fast development combined with simple sample preparation methods particularly for matrix-assisted laser desorption/ ionization MS (MALDI-MS). Madonna et al.22 first introduced a method of phage amplification coupled to MALDI-MS analysis to produce a charged molecular species signal for the

■

EXPERIMENTAL PROCEDURES

Chemicals and Reagents

Sequencing-grade modified trypsin was from Promega (Madison, WI). Synthetic peptides were partially synthesized by Bachem (Burgdorf, Switzerland) and in-house (bioMérieux, Marcy l’Etoile, France). Dithiothreitol (DTT), iodoacetamide (IAA), sodium chloride, gelatin, and analytical-grade formic acid (FA) were from Sigma-Aldrich (Sigma Chemical, St. Louis, MO). Ultrapure water was from a Milli-Qplus 185 purifier (Millipore, Bedford, MA), and HPLC-grade acetonitrile (ACN) was from SDS (Peypin, France). Cassoulet (French stew) was sampled in a commercial sterile can, and orange juice was purchased in a local supermarket. RapiGest SF (lyophilized sodium-3-[(2-methyl-2-undecyl-1,3-dioxolan-4-yl)-methoxyl]1-propanesulfonate) was from Waters (Milford, MA). Buffered peptone water (BPW) and bag-stomacher were from bioMérieux (Marcy l’Etoile, France). 1451

dx.doi.org/10.1021/pr400991t | J. Proteome Res. 2014, 13, 1450−1465

Journal of Proteome Research

Article

Bacterial and Phage Strains

Biomarker Discovery

E. coli (ATCC 11303) and phage T4 (ATCC 11303-B4) were obtained from the Félix d’Hérelle Reference Center for bacterial viruses of the Université Laval (GREB, QC, Canada). B. subtilis strain YB88638 and the phage SPP139 were from the Unité de Virologie Moléculaire et Structurale (VMS) strain collection. All microbiological procedures were performed in a biological safety laboratory level 2. Bacterial samples were cultured in Luria−Bertani (LB) broth (Difco, Detroit, MI). Bacterial cell concentrations were estimated using standard optical density (OD) measurements at 600 nm.

(i). Protein Extraction and Digestion. Bacterial cells were infected with phages using a multiplicity of infection (MOI; i.e., the number of phages per bacterium ratio) of 3 (E. coli/T4 model) and 5 (B. subtilis/SPP1 model). The resulting mixtures were then incubated with continuous shaking for 5 to 30 min (37 °C). In the case of phage T4, 1 mL aliquots were withdrawn every 5 min, mixed with 3 mL of cold LB, and centrifuged to sediment bacteria and eliminate free input phages that remain in the supernatant (10 min at 3900g at 4 °C). An identical procedure was followed for SPP1-infected bacteria with the exception that 2 mL aliquots were taken at each time point, sedimented for 3 min at 16 100g (room temperature) in a microfuge, followed by elimination of the supernatant and a short spin of the tube containing the pellet for careful pipetting away of the remaining supernatant. Bacterial pellets were then rapidly frozen at −80 °C until analysis. Protein extraction from whole bacterial cells was performed using ethanol−FA according to Freiwald and Sauer’s protocol.43 Extracts were then analyzed in parallel both for intact proteins and for peptides, resulting from trypsin digestion by LC−MS/MS. Solution digestion for peptide analysis was adapted from the protocol of Fenaille et al.44 In brief, 20 μL aliquots of the protein extract were solubilized with 10 μL of 4 mg/mL RapidGest solution, reduced by 5 μL of 100 mM DTT solution for 10 min at 95 °C, and alkylated by 5 μL of 240 mM IAA solution at room temperature for 45 min. Enzymatic digestion was performed by the addition of 1 μg of trypsin and overnight incubation at 37 °C. A 5 μL aliquot of 1 M HCl was added to cleave the acid-labile surfactant, and samples were incubated at 37 °C for 45 min and then centrifuged (10 min at 11 300g). Finally, the supernatant was recovered, and the digested samples were transferred to autosampler vials for LC− MS/MS analysis. For intact protein profiling, 20 μL aliquots of the protein extracts were transferred to autosampler vials for LC−MS/MS analysis. After profiling, discriminating proteins were manually collected from the liquid chromatography eluate and identified as described later in the Protein Identification section. (ii). LC−MS/MS Conditions. Twenty μL of sample (intact or digested proteins) was analyzed on a 2.1 × 150 mm Zorbax C18-300SB column (5 μm particle size) (Agilent Technologies) using an Accela liquid chromatography system (Thermo Fisher Scientific). The aqueous mobile phase (A) consisted of HPLC-grade water with 0.1% FA, while the organic phase (B) was ACN with 0.1% FA. A gradient profile was used at a flow rate of 200 μL/min. Initially the mobile phase consisted of 5% B and 95% A, ramped to 60% B over 40 min, continuing up to 95% B over the following 5 min and holding for 3 min. After a 50-min run time, the gradient was returned to 5% B for the next 10 min to equilibrate the column to initial conditions. Total run time was 60 min. The column temperature was maintained at 40 °C. The column eluent was introduced into a Thermo Scientific LTQ-Orbitrap Discovery mass spectrometer equipped with an electrospray interface (Thermo Scientific, San Jose, CA). Analyses were performed in the positive ion mode. The electrospray voltage, capillary voltage, and tube lens voltage were set to 5 kV, 35 V, and 70 V, respectively. The sheath, ion sweep, and auxiliary gas flows (nitrogen) were set at 35, 20, and 10 (arbitrary units), and the capillary temperature was set at 275 °C. For MS experiments, the resolving power of the

Phage Stock Preparation

(i). Phage T4. E. coli cells were grown overnight at 37 °C in LB broth with constant shaking (220 rpm). LB broth was then inoculated with 1:100 volumes of the overnight culture and incubated at 37 °C, with shaking (220 rpm) until the exponential growth phase was reached (OD = 0.4). Cells were then infected with phage T4 for 2 h with constant shaking. Phage lysates were cleared by centrifugation (3900g for 10 min at 4 °C), and viral particles were then sedimented by overnight centrifugation at 12 000 g (4 °C). Pellets were suspended in TSG buffer (0.01 M Tris-HCl pH 7.4, 0.15 M NaCl, 0.03% gelatin). A second centrifugation (10 min at 3900g) was used to obtain a clear phage suspension. The resulting stock solution of phage T4 was stored at 2−5 °C. Phage concentrations were determined using the standard double agar overlay plaque assay.40 For electron microscopy (EM) observation, the phage preparation was purified by centrifugation in a discontinuous CsCl gradient.41 Viral particles were then dialyzed against 100 mM Tris-Cl (pH 7.5)−10 mM MgCl2 using a Slide-A-Lyzer G2 dialysis cassette (3.5K MWCO, Thermo Scientific), with an NaCl gradient (from 3 to 0.1 M) as described.41 Phages were stored at 4 °C until preparation for EM by negative staining with ammonium molybdate 2%, followed by observation on a Phillips CM12 electron microscope by the IBS/UVHCI platform (Grenoble, France). (ii). Phage SPP1. Bacteriophage SPP1 was multiplied in B. subtilis YB886, sedimented, and purified through a discontinuous CsCl gradient as described.42 Amplification of Phage T4 in Different Media and Recovery

T4 was amplified in “cassoulet”, orange juice, and LB broth. For the solid matrix, 25 g of “cassoulet” was crushed in a 400 mL stomacher containing 225 mL of BPW. E. coli cells in exponential phase were grown to an OD600nm of 0.4 (108 CFU/mL) and spiked in the stomacher to a final concentration of 105 to 106 CFU/mL. For the liquid matrix, 2 mL of orange juice was diluted 1:5 in 10 mL of buffered LB containing ammonium bicarbonate 0.4 M to raise the pH to 7. E. coli cells were spiked in the diluted juice for a final concentration of 105 to 107 CFU/mL. In LB broth, E. coli cells were spiked for a final concentration of 104 to 108 CFU/mL. All experiments were carried out in duplicate. The E. coli cells in the different media were then infected with a constant input of phage (