Detection of Yersinia pestis in Environmental and Food Samples by

May 21, 2014 - *Phone: 33-1-69-08-13-15. .... Detection of Yersinia pestis in Complex Matrices by Intact Cell Immunocapture and Targeted Mass Spectrom...
0 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/ac

Detection of Yersinia pestis in Environmental and Food Samples by Intact Cell Immunocapture and Liquid Chromatography−Tandem Mass Spectrometry Jérôme Chenau,†,‡ François Fenaille,† Stéphanie Simon,† Sofia Filali,‡ Hervé Volland,† Christophe Junot,† Elisabeth Carniel,‡ and François Becher*,† †

Service de Pharmacologie et d’Immunoanalyse, Institut de Biologie et de Technologies de Saclay (iBiTec-S), Commissariat à l’Énergie Atomique et aux Énergies Alternatives (CEA), 91191 Gif-sur-Yvette, France ‡ Unité de Recherche Yersinia, Institut Pasteur, 75724 Paris Cedex 15, France S Supporting Information *

ABSTRACT: Yersinia pestis is the causative agent of bubonic and pneumonic plague, an acute and often fatal disease in humans. In addition to the risk of natural exposure to plague, there is also the threat of a bioterrorist act, leading to the deliberate spread of the bacteria in the environment or food. We report here an immuno-liquid chromatography−tandem mass spectrometry (immuno-LC−MS/MS) method for the direct (i.e., without prior culture), sensitive, and specific detection of Y. pestis in such complex samples. In the first step, a bottom-up proteomics approach highlighted three relevant protein markers encoded by the Y. pestis-specific plasmids pFra (murine toxin) and pPla (plasminogen activator and pesticin). Suitable proteotypic peptides were thoroughly selected to monitor the three protein markers by targeted MS using the selected reaction monitoring (SRM) mode. Immunocapture conditions were optimized for the isolation and concentration of intact bacterial cells from complex samples. The immuno-LC−SRM assay has a limit of detection of 2 × 104 CFU/mL in milk or tap water, which compares well with those of state-of-the-art immunoassays. Moreover, we report the first direct detection of Y. pestis in soil, which could be extremely useful in confirming Y. pestis persistence in the ground.

pMT1) and pPla (or pPst or pPCP1), which code for virulence factors representing interesting targets for specific detection.1,18 Nucleic acid-based assays12,19,20 and immunometric tests21 have been developed for the direct detection of Y. pestis by targeting genes or proteins coded by the specific pFra and pPla plasmids. However, the complexity of environmental and food matrixes may interfere with bacterial detection due to inhibition of PCR amplification22,23 and/or antibody cross-reactions.24 Additionally, reference immunoassays used in plague diagnosis are based on the detection of the pFra-encoded F1 antigen.25 They cannot be applied to environmental or food materials since F1 antigen expression is drastically decreased (about 40− 100-fold in gene expression) at temperatures below (26−28 °C) the body temperature of 37 °C.1,26,27 Mass-spectrometry-based proteomics is particularly well suited for biodefense applications and detection of pathogenic microorganisms.28 The reproducible acquisition of global bacterial protein fingerprints/patterns by matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) has been applied to species differentiation within the genus Yersinia.29,30 However, specific discrimination

Yersinia pestis, a highly virulent bacterium, is the etiologic agent of bubonic and pneumonic plague, an acute and often lethal infection of animals and humans.1 Throughout recorded history, the world has witnessed three pandemics of plague that caused about 200 million human deaths.1 Nowadays, natural plague foci still persist in Asia, Africa, and the Americas, and about 2000 cases of human plague are recorded by the World Health Organization every year.2,3 Due to its prevalence, high lethality, and easy person-to-person dissemination,4−6 Y. pestis is classified as a category A biothreat agent by the Centers for Disease Control and Prevention (CDC), the highest rank of potential bioterrorism agents (http://www.bt.cdc.gov/agent/ agentlist-category.asp). Given the reported persistence of Y. pestis in the environment (soil7−9 and water10) and the threat of foodborne bioterrorism,11,12 efficient methods for detection and confirmation of the presence of bacteria in these matrixes are required. Detection of Y. pestis is complicated by the high genetic similarity of the populations composing the Yersinia pseudotuberculosis complex (Y. pestis, Y. pseudotuberculosis, Yersinia similis, and Yersinia wautersii).13,14 Y. pestis emerged approximately 2500 years ago from Y. pseudotuberculosis,15,16 and the two species have ∼97% nucleotide identity.17 Y. pestis is mainly characterized by the acquisition of two plasmids, pFra (or © 2014 American Chemical Society

Received: April 15, 2014 Accepted: May 21, 2014 Published: May 21, 2014 6144

dx.doi.org/10.1021/ac501371r | Anal. Chem. 2014, 86, 6144−6152

Analytical Chemistry

Article

Sample Preparation. Magnetic beads (Dynabeads M-280 tosyl-activated, Life Technologies, Carlsbad, CA) were prepared by following the protocol provided by the supplier. Classically, monoclonal antibody Pla35 (220 μg) was covalently linked to the beads to obtain 500 μL of IgG-coupled bead solution. IgG-coupled bead solution (35 μL) was added to 1 mL of bacteria diluted in PBS containing 0.1% Tween-20 and 1 mg/mL BSA (ratio 1:1, v/v). Samples were incubated for 90 min at room temperature with gentle shaking and then washed twice with PBS to remove weak nonspecific binding. Proteins were extracted from the immunocaptured bacteria directly on beads by using the TFA protocol for highly pathogenic microorganisms described by Lasch et al.36 Briefly, 150 μL of 80% TFA solution was added to the beads−IgG−bacteria complex, and the mixture was vortexed for 10 min. Beads were then retained on the magnetic support, and the supernatant was centrifuged at 14000g and 4 °C for 15 min. Finally, the supernatant was transferred into Ultrafree MC filter tubes of 0.22 μm pore size (Millipore, Billerica, MA) and spun at 10000g for 5 min. The filtrate containing the total protein extract was stored at −20 °C until use. The total inactivation and elimination of bacteria were routinely checked by verification of the absence of colonies after culture on LBH agar as described previously.36 Each protein extract was dried by vacuum centrifugation to eliminate TFA. Proteins were then resuspended in 25 μL of 50 mM ammonium bicarbonate (NH4HCO3) buffer, pH 8.0. Each sample was reduced by the addition of 5 μL of 45 mM dithiothreitol (DTT) solution (in 50 mM NH4HCO3 buffer) at 60 °C for 30 min and alkylated by the addition of 5 μL of 100 mM iodoacetamide (IAA) solution (in 50 mM NH4HCO3 buffer) at room temperature for 45 min. Finally, enzymatic digestion was performed by the addition of 1.5 μL of 1 μg/μL trypsin or Glu-C with overnight incubation at 37 °C. Bottom-Up Proteomics Approach. Enzymatically digested samples were diluted in 0.1% formic acid (1:5, v/v) and analyzed in triplicate by liquid chromatography−tandem mass spectrometry (LC−MS/MS) using a fast LC gradient going from 5% to 60% acetonitrile in only 40 min and a datadependent “top 5” MS method. LC−MS/MS experiments were performed using an Accela HPLC system coupled to an LTQOrbitrap Discovery mass spectrometer, both from Thermo Scientific (San Jose, CA). Chromatographic separation was performed on a Zorbax SB-C18 column (150 × 2.1 mm i.d., 5 μm particle size, 300 Å porosity) from Agilent Technologies (Palo Alto, CA). The detailed chromatographic conditions and MS parameters are described in the Supporting Information. Raw files were converted to Mascot compatible .mgf files using Proteome Discoverer version 1.3 (Thermo Scientific). Database searches were performed using Mascot server version 2.3.01 through Proteinscape platform version 3.0 (Bruker Daltonics, Bremen, Germany) using the protein database Y. pestis [taxonomy 632] (93066 entries) downloaded from UniProtKB/TrEMBL (updated Dec 13, 2011). The parameters for searching were trypsin or Glu-C (V8-DE) as the digestion enzyme, two missed cleavages allowed, charge states 2+ and 3+, fixed modification of cysteine residues (carbamidomethyl), variable oxidation of methionine residues, deamidation of asparagine and glutamine residues, Gln → pyro-Glu (for Nterminal glutamine residues) and loss of N-terminal methionine (−131.04 Da), parent ion tolerance of 10 ppm, and fragment mass tolerance of 0.7 amu. Protein identifications were

of the different species composing the Y. pseudotuberculosis complex is still challenging.14,29 Moreover, such an untargeted approach requires bacterial cultivation and therefore leads to a long time to results since Y. pestis needs incubation periods of 36−48 h.31,32 Up to now, and to the best of our knowledge, no MS approach has been developed for direct and specific detection of Y. pestis in complex matrixes. Such a method would be useful in the context of a bioterrorism event for which a combination of complementary methods is necessary for reliable detection. Targeted proteomics based on the selected reaction monitoring (SRM) mode offers an attractive approach for the sensitive, multiplex, and accurate detection and/or quantification of proteins.33 Regarding detection in complex matrixes, specific isolation and concentration of the targeted component could significantly enhance mass spectrometry detection. In that respect, immunocapture has been shown to be an efficient tool in enriching various bioterrorism-related agents such as vegetative cells,34 toxins,35and spores.23 In this study, we report the first immuno-LC−SRM approach for the sensitive and specific detection of Y. pestis in complex environmental and food samples. In the preliminary step, a bottom-up proteomics approach was used to highlight three relevant protein markers of Y. pestis to be targeted by SRM. The whole assay relies on the immunocapture of intact bacteria, followed by protein extraction, proteolysis, and a multiplex SRM detection of the proteotypic peptides. It allows detection of Y. pestis in complex matrixes at levels as low as 2 × 104 CFU/ mL.

■

EXPERIMENTAL SECTION Chemicals and Reagents. Luria−Bertani (LB) agar was from Difco (Detroit, MI), hemin from Acros Organics (Geel, Belgium), and phosphate-buffered saline (PBS) from Lonza (DPBS without Ca2+ and Mg2+, Verviers, Belgium). Sequencing-grade modified trypsin was from Promega (Fitchburg, WI) and endoproteinase Glu-C from Sigma-Aldrich (St. Louis, MO). Synthetic isotopically labeled peptides were synthesized by Bachem (Burgdorf, Switzerland). Mouse monoclonal antibody Pla35 was directed against Y. pestis surface protein plasminogen activator (Pla) and produced as previously described.21 Sodium phosphate monobasic (NaH2PO4) and dibasic (Na2HPO4), Tween-20, and BSA were from SigmaAldrich. Ultrapure water was from a Milli-Qplus 185 purifier (Millipore, Bedford, MA), HPLC-grade acetonitrile (ACN) was from SDS (Peypin, France), and analytical-grade formic acid and trifluoroacetic acid (TFA) were from Sigma-Aldrich. Commercial milk powder was prepared at 5% (w/v) in ultrapure water. Soil was collected in a field around the laboratory. Bacterial Strains and Culture Conditions. The Y. pestis (n = 7), Y. pseudotuberculosis (n = 8), and Y. similis (n = 2) isolates used in this study are listed in Table S-1 (Supporting Information). As Y. pestis is a highly virulent species, strains were handled according to safety considerations applicable at the Institut Pasteur (Biosafety Level 3 laboratory). Bacteria were grown at 28 °C (and at 37 °C for some controls) on Luria−Bertani agar plates supplemented with 0.2% hemin (LBH) for 48 h before use. Bacteria were resuspended in sterile saline solution (0.9% NaCl), and bacterial concentrations were evaluated by OD measurement at 600 nm and plating on LBH plates. 6145

dx.doi.org/10.1021/ac501371r | Anal. Chem. 2014, 86, 6144−6152

plasminogen activator protease (Pla) pesticin (Psn) murine toxin (Ymt) F1 capsule antigen (F1) F1 chaperone protein (Caf1M) F1 capsule anchoring protein (Caf1A) outer membrane protein A (OmpA)

plasminogen activator protease (Pla) pesticin (Psn) murine toxin (Ymt) F1 capsule antigen (F1) F1 ch aperone protein (Caf1M) F1 capsule anchoring protein (Caf1A) outer membrane protein A (OmpA)

D0JML5 D0JML3 D0JMJ7 D0JMI9 D0JYB6 D0JMJ0 D0JJ85

D0JML5 D0JML3 D0JMJ7 D0JMI9 D0JYB6 D0JMJ0 D0JJ85

34611 40043 67547 17666 28751 89734 37930

34611 40043 67547 17666 28751 89734 37930

MW

823.2 687.1 964.7 ND ND ND 1006.5

966.2 481.3 823.7 882.5 596.5 110.8 989.6

score

14 12 22 ND ND ND 21

17 9 19 11 12 2 21

no. of peptides

49.7 41.7 44.3 ND ND ND 57.8

50.3 34.5 45.5 65.3 50.0 3.2 60.3

coverage (%) 445.2 481.3 180.7 497.9 84.8 ND 989.6

174.5 220.0 215.9 ND ND ND 666.8

28 °C 0.88 0.88 1.11 ND ND ND 0.73

score

37 °C 1.36 0.79 2.04 1.23 0.96 0.87 1.27

RMS90 (ppm)

2 6 5 ND ND ND 13

9 9 4 9 2 ND 21

no. of peptides

9.3 10.6 9.2 ND ND ND 43.3

35.4 34.5 8.0 52.9 9.3 ND 60.3

coverage (%)

Glu-C

0.87 0.75 1.16 ND ND ND 1.15

1.03 0.79 1.50 1.13 0.60 ND 1.27

RMS90 (ppm)

16 18 27 ND ND ND 34

21 13 23 20 14 2 30

no. of peptides

total

55.1 47.3 45.7 ND ND ND 73.9

56.1 40.6 46.7 82.4 53.1 3.2 61.5

coverage (%)

pPla plasmid pPla plasmid pFra plasmid pFra plasmid pFra plasmid pFra plasmid chromosome

pPla plasmid pPla plasmid pFra plasmid pFra plasmid pFra plasmid pFra plasmid chromosome

localization

For each protein, the UniProt accession number and molecular weight (MW) are represented. Bottom-up experiments were performed using trypsin and Glu-C enzymes. The mascot protein score (score), number of peptides matched (no. of peptides), protein sequence coverage (coverage, %), and root-mean-square (RMS90, ppm) are shown. ND = not detected.

a

protein

accession no.

trypsin

Table 1. Protein Markers Identified by the Bottom-Up Approach for Y. pestis CO92 Grown at 37 and 28 °Ca

Analytical Chemistry Article

6146

dx.doi.org/10.1021/ac501371r | Anal. Chem. 2014, 86, 6144−6152

Analytical Chemistry

Article

validated if proteins were identified with one peptide with a score >40 or with two peptides with a score >20. Multiplex SRM Assay. Stable-isotope-labeled peptides corresponding to each targeted marker were synthesized to be used as internal standards and to develop the assay (Table S2, Supporting Information). The preparation of the internal standard mixture is described in the Supporting Information. LC−MS/MS experiments were performed using an HP 1100 HPLC system from Agilent coupled to a triple-quadrupole TSQ Quantum Ultra mass spectrometer (Thermo Scientific). Chromatographic separation was performed on a Zorbax SBC18 column (150 × 2.1 mm i.d., 5 μm particle size, 300 Å porosity) from Agilent. The detailed chromatographic conditions and MS parameters are described in the Supporting Information. The LC−SRM method takes 30 min. Several SRM transitions were monitored for each peptide to improve detection specificity. Product ions were preferentially selected with an m/z ratio higher than that of the parent ion. A signal was considered positive when peak intensity was 3 times above the noise level for selected transitions. The SRM acquisition time was divided into five segments. For a given peptide, the retention time observed for the transitions monitored and the ratio between their peak areas must match those obtained with their corresponding internal standard peptides. In Table S-2 we report the SRM transitions designed and monitored for each peptide and the optimized analytical parameters (tube lens and collision energy values).

showing higher protein content at this optimal growth temperature.39 Among them, the F1 antigen (F1), encoded by the pFraborne gene caf1, was identified only at 37 °C (Table 1, upper part), which corroborated previous data at the gene17,26 and protein37 levels. The F1 protein was abundantly produced at this temperature, as may be reflected by its high Mascot score, the 20 unique peptides identified upon the database search leading to a high protein sequence coverage (82.4%, corresponding to 95.3% of the mature F1 protein) (Table 1, upper part). The Caf1M and Caf1A proteins, involved in the assembly of F1 subunits at the surface of the bacteria, were also identified at 37 °C only (Table 1), which precludes their use for Y. pestis detection in environmental and food samples. Interestingly, three other markers encoded by the Y. pestisspecific plasmids were identified at both temperatures: the murine toxin (Ymt) encoded by pFra and the plasminogen activator (Pla) and pesticin (Psn) encoded by pPla (Table 1). Blast similarity searches against the whole UniProt database confirmed the high sequence specificity of each of them for Y. pestis. These proteins are among the best hits at both temperatures, with high numbers of unique peptides matched to them and high protein sequence coverages (Table 1; Table S-3, Supporting Information). Altogether these data demonstrate that such abundant proteins could constitute particularly interesting targets for the sensitive and broad-spectrum detection of Y. pestis. It should be noted that a recent study has pointed to the hypothetical protein YPO1670 as a putative chromosome-encoded marker to target for the identification of Y. pestis.40 However, this putative protein was not identified in our conditions (Table S-3). Moreover, no other chromosomeencoded proteins have been highlighted as specific potential markers in a first screening approach. Ymt is a phospholipase D, which is involved in Y. pestis transmission by promoting its survival in and colonization of the flea midgut.41 Psn is a bacteriocin, which kills surrounding bacteria.42 Pla is an outer membrane protease belonging to the omptin family which cleaves mammalian plasminogen and plays a key role in Y. pestis virulence, as its inactivation increases the median lethal dose of the bacteria one million-fold.43,44 These Ymt, Pla, and Psn proteins were selected for the multiplex targeted detection of Y. pestis. In addition, outer membrane protein A (OmpA) was selected to evaluate the nonspecific capture of the closely related Y. pseudotuberculosis and Y. similis species, which share a common isoform of this protein (Table 1; Table S-3, Supporting Information). Analytical Strategy for Targeted Detection of Y. pestis. The overall targeted immuno-LC−MS/MS (SRM) procedure developed for sensitive multiplex detection of Y. pestis and its discrimination from other closely related species (Y. pseudotuberculosis and Y. similis) is schematized in Figure S-1 (Supporting Information). Development of the LC−SRM Method. The LC−SRM method had to monitor specifically the three Y. pestis-specific proteins Pla, Ymt, and Psn, plus OmpA, which is shared by the different members of the Y. pseudotuberculosis complex. Indepth analysis of proteomics data was performed to select the most suitable peptides for robust and sensitive LC−SRM detection. Proteotypic peptides were selected to obtain 8−25 amino acid lengths, a good chromatographic behavior, and a high ionization efficiency, while avoiding methionine, cysteine, or tryptophan residues that could undergo oxidation during sample handling. Peptides with two neighboring basic amino

■

RESULTS AND DISCUSSION Selection of Y. pestis-Specific Markers Using a Bottom-Up Proteomics Approach. With the aim of developing a targeted MS method for the robust detection of Y. pestis, a prior selection of specific markers was required. Proteins coded by pPla and pFra plasmids are unique to Y. pestis and represent suitable targets for its specific detection.1,18 We therefore performed a bottom-up proteomics approach to determine which proteins coded by these plasmids could be detected among the most abundant ones of the Y. pestis proteome. Moreover, as gene expression and protein synthesis are modulated by temperature in Y. pestis,26,37 proteomics analysis was performed on bacteria grown at either 28 °C (optimal growth temperature) or 37 °C (temperature during infection)1 to select markers detectable at both temperatures. This proteomics analysis was also useful to highlight the most suitable proteotypic peptides for the subsequent specific, sensitive, and robust Y. pestis detection. Protein extracts from Y. pestis CO92 were subjected to either trypsin or Glu-C digestion. The use of two enzymes with different specificities has been shown to increase the number of peptides/proteins identified.38 The resulting peptide mixture was analyzed in triplicate by LC−MS/MS (using an LTQOrbitrap instrument) in a shotgun manner oriented toward the detection of major Y. pestis proteins. The MS/MS data were subjected to a Mascot database search and led to the identification of 3255 peptides at 28 °C (2281 tryptic and 974 Glu-C peptides) and 3122 peptides at 37 °C (2208 tryptic and 914 Glu-C peptides), with an average mass accuracy better than 2 ppm for the parent ions. These peptide numbers further correspond to 515 and 445 distinct proteins at 28 and 37 °C, respectively (Table S-3, Supporting Information). The largest number of proteins observed at the optimal growth temperature of 28 °C may be correlated with recent observations 6147

dx.doi.org/10.1021/ac501371r | Anal. Chem. 2014, 86, 6144−6152

Analytical Chemistry

Article

Figure 1. Typical LC−SRM chromatograms obtained for (A) Y. pestis CO92 and (B) Y. pseudotuberculosis IP32953. Depicted SRM transitions are as follows: Pla (peptide 1 NSGDSVSIGGDAAGISNK, 825 → 1089; peptide 2 INDFELNALFK, 662.5 → 834.5), Psn (ENEDLNNNR, 615.6 → 630.3), Ymt (peptide 1 ILIAPFFFTDK, 656.5 → 972.5; peptide 2 EIMQQSYLR, 584.1 → 666.4), and OmpA (peptide 1 LSYPVAQDLDVYTR, 820.4 → 638.8; peptide 2 GSFDGGLDR, 462.2 → 517.3). In each case, the signal of the corresponding isotopically labeled peptide is indicated by IS (internal standard). Data were obtained using 108 bacteria.

acids or RP/KP sequences were also excluded.45 Proteins sharing homologies with Pla have been described in other species (e.g., Klebsiella pneumoniae, Salmonella enterica, or Erwinia pyrifoliae),46 which, unlike in Y. pestis, are chromosomeencoded. Therefore, special care was taken to select proteotypic peptides specifically belonging to the Y. pestis Pla proteoform. This was further confirmed by performing Blast similarity searches against the whole UniProt database. Two proteotypic peptides were selected for each protein, except for Psn, for which only one adequate peptide could be retrieved. Overall, seven proteotypic peptides were selected to specifically monitor the four proteins (Table S-2, Supporting Information). Although peptide 2 selected for Ymt contained a methionine residue, the presence of the oxidized form was consistently below 10% in our proteomics data (data not shown). This peptide was therefore only used for qualitative analysis in complement to peptide 1. The seven peptides were produced in an isotopically labeled form to be used as internal standards and to optimize MS and MS/MS conditions. Several SRM transitions were monitored per peptide for improved specificity. Moreover, product ions were preferentially selected with an m/z ratio higher than that of the parent ion. Coelution of the different SRM traces and matching between area ratios are mandatory for confident peptide identification and quantification. In Table S-2 (Supporting Information) we report the SRM transitions designed and monitored for each peptide and the optimized analytical parameters (tube lens and collision energy values). The final LC−MS/MS method monitored 34 different SRM transitions in 30 min, corresponding to the four targeted proteins (Figure S-2, Supporting Information). Typical LC− MS/MS chromatograms obtained with Y. pestis CO92 and the closely related Y. pseudotuberculosis IP32953 are shown in Figure 1. As expected, positive signals corresponding to the three markers Pla, Psn, and Ymt and OmpA were observed for

Y. pestis (Figure 1 A). Conversely, no signal corresponding to the three Y. pestis-specific markers was observed in the closely related Y. pseudotuberculosis strain (Figure 1 B), while OmpA was still detected. Development of Intact Cell Immunocapture. Application to complex samples requires isolation and concentration of Y. pestis prior to mass spectrometry analysis. For this aim, an immunocapture step was implemented to specifically retain intact Y. pestis cells (Figure S-1, Supporting Information). Pla is a suitable target for intact cell immunocapture, as it is specific to Y. pestis and is located at the bacterial surface.47 Various parameters were investigated to obtain optimal capture conditions. Magnetic beads with two different diameters (300 nm and 2.8 μm) and chemistries for antibody coating (covalent binding with tosyl-activated beads or by protein G affinity) as well as different homemade monoclonal antibodies directed against the specific membrane protein Pla21 were tested. These antibodies have previously been evaluated for Y. pestis immunodetection.21 Optimal recovery was achieved using tosyl-activated magnetic beads (2.8 μm diameter) coated with Pla35 antibody. The dilution of bacterial samples in PBS buffer containing 0.1% Tween-20 and 1 mg/mL BSA and incubation for 90 min at room temperature with gentle shaking on a wheel were required for a high and specific capture yield, weak nonspecific binding on the magnetic bead surface or wall tubes, and full integrity and low autoaggregation of Y. pestis cells.48 OmpA proteotypic peptides were monitored to evaluate the efficiency of the Pla-dependent immunocapture of Y. pestis and its specificity toward the closely related Y. pseudotuberculosis. The immunocapture yield was calculated as the percentage of the normalized SRM areas observed with and without immunocapture. Under optimal conditions, immunocapture of Y. pestis CO92 proved to be efficient with a yield of more than 75% (Figure S-3, Supporting Information). Conversely, 6148

dx.doi.org/10.1021/ac501371r | Anal. Chem. 2014, 86, 6144−6152

Analytical Chemistry

Article

Table 2. Reproducibility of the Immunocapture Step and the Whole Immuno-LC−MS/MS (SRM) Assaya Pla IC yield immuno-LC−MS/MS assay

Psn

Ymt

OmpA

Pep1

Pep2

Pep1

Pep1

Pep2

Pep1

Pep2

av

70.1 7.1 9.1

79.3 4.8 6.2

70.6 5.7 8.2

70.9 4.1 6.6

78.7 5.9 6.7

73.2 4.6 5.1

85.6 4.8 5.4

75.5 5.3 6.8

av value (%) std dev (%) CV (%)

Data were obtained by independently preparing and analyzing nine replicates of Y. pestis CO92 (1 mL at 5 × 107 CFU/mL in buffer). For each peptide, the average immunocapture yield (av value, %) and the corresponding interassay standard deviation (std dev, %) were calculated. The coefficient of variation (CV, %) of the normalized peak area obtained, with the whole immuno-LC−MS/MS assay, for each targeted peptide is also indicated.

a

the immunocapture yield with Y. pseudotuberculosis IP32953 was below 1%, indicating a very low level of capture. The reproducibility of both the immunocapture step and the whole immuno-LC−MS/MS (SRM) assay were evaluated by independently preparing and analyzing nine replicates of Y. pestis CO92 (Table 2). The average immunocapture yields measured for each monitored peptide were similar, in the range of 75%, with low variation between the replicates (interassay standard deviation of 5.3%) (Table 2). Importantly, such high data consistency over the different markers correlated with the immunocapture of whole and intact bacterial cells through Pla. In the second step, the whole immuno-LC−MS/MS assay also proved to be reproducible, with a coefficient of variation (CV) of the normalized area from each targeted peptide of 6.8%, demonstrating that the whole assay has a good precision and reproducibility despite the multiple steps involved (i.e., intact cell immunocapture, protein extraction, enzymatic digestion, and LC−MS/MS analysis) (Table 2). Applications of the Immuno-LC−SRM Assay. Detection Specificity for Y. pestis. Specificity is crucial for Y. pestis detection. To establish the specificity of the whole assay, 17 different strains were analyzed (Table S-1, Supporting Information). Seven Y. pestis strains belonging to the three most common biovars (Antiqua, Medievalis, and Orientalis) and isolated from diverse continents (Asia, Africa, and South and North America) (Table S-1) were selected. Ten other strains from the Y. pseudotuberculosis complex (eight Y. pseudotuberculosis strains of the most common serotypes and two Y. similis strains) were also included and analyzed as controls (Table S-1). The immuno-LC−MS/MS (SRM) assay has two levels of specificity. The first level was achieved by the specific antibody capture. The immunocapture yield was between 55% and 100% for Y. pestis (mean of 77%) and about 2% for the other strains of the Y. pseudotuberculosis complex (Table 3; Table S-4,

Supporting Information). The second level of specificity was provided by the SRM analysis itself. Of the 17 strains screened, the 7 Y. pestis strains gave positive signals for all three Y. pestis markers (Pla, Psn, and Ymt), whereas the 10 other strains of the Y. pseudotuberculosis complex did not yield any signal, demonstrating the specificity of the assay (Tables 3 and S-4). We also tested the efficiency of the method in detecting Y. pestis grown at the body temperature of 37 °C. Our proteomics data showed that Pla, Psn, and Ymt are produced at both 28 and 37 °C. The SRM experiments confirm these results (Figure S-4, Supporting Information). As previously reported with ELISA and dipstick methods using the same Pla35 monoclonal antibody,21 Y. pestis cells grown at 37 °C can also be captured. The immunocapture yield was significantly lower at 37 °C (35%) than at 28 °C (77%) (data not shown), maybe because the high abundance of F1 antigen at the surface of the bacteria at 37 °C partly inhibits the accessibility of the antibodies to Pla. Applications to the Detection of Y. pestis in Environmental Samples. In the case of a suspected bioterrorism event, it is essential to ascertain whether environmental samples such as food, water, or aerosol droplets are contaminated with Y. pestis. Moreover, survival of Y. pestis in soil and rodent burrows under experimental or natural conditions has previously been demonstrated,7−9,49 which makes these matrixes also pertinent for screening. To validate the applicability of the immuno-LC−MS/MS method to these samples and to evaluate the resulting sensitivity, we analyzed the detection efficiency in three complex matrixes: tap water, milk, and soil. In contrast to that in tap water, detection of Y. pestis in milk and soil was rather unsuccessful without prior immunocapture, as previously observed for Bacillus anthracis spores.23 For tap water and milk samples, different quantities of the Y. pestis CO92 strain were spiked into 10 mL of matrix. The bacteria were easily extracted and enriched from such samples by the immunocapture step. At the limit of detection, the monitoring of the most intense Pla and Ymt peptides was sufficient to have a robust, sensitive, and specific detection of Y. pestis. The limit of detection (LOD), based on these criteria, was 2 × 104 CFU/mL in milk and tap water (Table 4). Figure 2 shows examples of regression curves obtained for the detection of Y. pestis in milk and the corresponding LC−MS/MS chromatograms for an LOD of 2 × 104 CFU/mL. The coefficient of correlation (r2) indicated a good linearity over a 4-log concentration range (2 × 104 to 2 × 108 CFU/mL), demonstrating that matrix components do not significantly impact the robustness and sensitivity of the method whatever the concentration. Moreover, tap water and milk blank samples did not exhibit any interfering signal, thus underlining the efficiency of the immunocapture and the specificity of the LC− MS/MS detection.

Table 3. Specificity of Detection for 7 Y. pestis Strains Compared with 10 Non-Y. pestis Strains of the Y. pseudotuberculosis Complexa IC yieldb (%)

SRM detection

species

range

mean

Pla

Psn

Ymt

Y. pestis (n = 7) Y. pseudotuberculosis (n = 8) Y. similis (n = 2)

55−100 0.02−7.4 0.04−0.09

77.0 1.8 0.01

+ − −

+ − −

+ − −

a Experiments were performed in buffer. IC yield = immunocapture yield calculated as the percentage of the normalized peak area of targeted peptides with and without bacterial immunocapture. bFor Y. pseudotuberculosis and Y. similis strains, the IC yield was based on the monitoring of the two OmpA peptides.

6149

dx.doi.org/10.1021/ac501371r | Anal. Chem. 2014, 86, 6144−6152

Analytical Chemistry

Article

Table 4. Limit of Detection (LOD) of the Immuno-LC−SRM Assay in Different Matrixesa LOD Pla sample

unit

buffer tap water milk soil

CFU/mL CFU/mL CFU/mL CFU/10 mg

Pep1 2.0 1.8 2.0 9.0

× × × ×

104 104 105 105

Psn Pep2 2.0 9.0 2.0 4.5

× × × ×

Pep1

104 103 104 105

2.0 9.0 2.0 9.0

× × × ×

105 104 105 105

Ymt Pep1 2.0 1.8 2.0 4.5

× × × ×

103 104 104 105

Pep2 2.0 9.0 2.0 4.5

× × × ×

105 104 105 106

Pla + Ymt 2.0 1.8 2.0 4.5

× × × ×

104 104 104 105

a

The LODs are indicated for each monitored peptide of Pla, Psn, and Ymt. Peptide 2 of Pla and peptide 1 of Ymt are used for sensitivity. The LODs based on the detection of these two peptides are given in the column marked “Pla + Ymt”.

Figure 2. Assay sensitivity in milk. Milk samples (10 mL) were spiked with Y. pestis CO92 and analyzed following the process shown in Figure S-1 (Supporting Information). (A) Regression curves obtained in milk and noted as the number of initial bacteria spiked per milliliter (CFU/mL). Example for Pla (peptide 2) and Ymt (peptide 1). (B) LC−MS/MS chromatograms at the instrumental LOD, i.e., 2 × 104 CFU/mL. Depicted SRM transitions are as follows: Pla peptide 2 (transition 662.5 → 834.5) and Ymt peptide 1 (transition 656.5 → 972.5). In each case, the signal of the corresponding isotopically labeled peptide is indicated by IS (internal standard).

therefore evaluated the efficiency of immuno-LC−SRM in detecting Y. pestis in the presence of another bacterial species. For this purpose, mixtures composed of various concentrations of Y. pestis and Escherichia coli were prepared. Although an impact on the immunocapture efficiency was observed, the three markers were still detected, whatever the relative abundance of Y. pestis versus E. coli (as illustrated for Pla peptide 2 in Figure S-5, Supporting Information).

Soil samples are more challenging because of their solid, complex, and nonsterile characteristics. The LOD of the method was slightly impacted in this matrix, with 4.5 × 105 CFU detected starting from 10 mg of soil (Table 4). This lower LOD was due to a lower recovery rate of Y. pestis during the immunocapture step, which may in part be attributed to the presence of other microorganisms (bacteria, yeast, or fungi) in the soil. Indeed, a polycontamination of the soil samples was observed upon streaking on agar plates (data not shown). We 6150

dx.doi.org/10.1021/ac501371r | Anal. Chem. 2014, 86, 6144−6152

Analytical Chemistry

Article

(7) Eisen, R. J.; Petersen, J. M.; Higgins, C. L.; Wong, D.; Levy, C. E.; Mead, P. S.; Schriefer, M. E.; Griffith, K. S.; Gage, K. L.; Beard, C. B. Emerging Infect. Dis. 2008, 14, 941−43. (8) Mollaret, H. H. Bull. Soc. Pathol. Exot. Ses Fil. 1963, 56, 1168−82. (9) Yersin, A. Rev. Med. Suisse Romande 1994, 114, 393−95. (10) Torosian, S. D.; Regan, P. M.; Taylor, M. A.; Margolin, A. Can. J. Microbiol. 2009, 55, 1125−29. (11) Amoako, K. K.; Goji, N.; Macmillan, T.; Said, K. B.; Druhan, S.; Tanaka, E.; Thomas, E. G. J. Food Prot. 2010, 73, 18−25. (12) Amoako, K. K.; Shields, M. J.; Goji, N.; Paquet, C.; Thomas, M. C.; Janzen, T. W.; Bin Kingombe, C. I.; Kell, A. J.; Hahn, K. R. J. Pathog. 2012, 2012, 781652. (13) Laukkanen-Ninios, R.; Didelot, X.; Jolley, K. A.; Morelli, G.; Sangal, V.; Kristo, P.; Brehony, C.; Imori, P. F.; Fukushima, H.; Siitonen, A.; Tseneva, G.; Voskressenskaya, E.; Falcao, J. P.; Korkeala, H.; Maiden, M. C.; Mazzoni, C.; Carniel, E.; Skurnik, M.; Achtman, M. Environ. Microbiol. 2011, 13, 3114−27. (14) Savin, C.; Martin, L.; Bouchier, C.; Filali, S.; Chenau, J.; Zhou, Z.; Becher, F.; Fukushima, H.; Thomson, N. R.; Scholz, H. C.; Carniel, E. Int. J. Med. Microbiol. 2014, 304, 452−63. (15) Morelli, G.; Song, Y.; Mazzoni, C. J.; Eppinger, M.; Roumagnac, P.; Wagner, D. M.; Feldkamp, M.; Kusecek, B.; Vogler, A. J.; Li, Y.; Cui, Y.; Thomson, N. R.; Jombart, T.; Leblois, R.; Lichtner, P.; Rahalison, L.; Petersen, J. M.; Balloux, F.; Keim, P.; Wirth, T.; Ravel, J.; Yang, R.; Carniel, E.; Achtman, M. Nat. Genet. 2010, 42, 1140−43. (16) Cui, Y.; Yu, C.; Yan, Y.; Li, D.; Li, Y.; Jombart, T.; Weinert, L. A.; Wang, Z.; Guo, Z.; Xu, L.; Zhang, Y.; Zheng, H.; Qin, N.; Xiao, X.; Wu, M.; Wang, X.; Zhou, D.; Qi, Z.; Du, Z.; Wu, H.; Yang, X.; Cao, H.; Wang, H.; Wang, J.; Yao, S.; Rakin, A.; Li, Y.; Falush, D.; Balloux, F.; Achtman, M.; Song, Y.; Wang, J.; Yang, R. Proc. Natl. Acad. Sci. U.S.A. 2013, 110, 577−82. (17) Chain, P. S.; Carniel, E.; Larimer, F. W.; Lamerdin, J.; Stoutland, P. O.; Regala, W. M.; Georgescu, A. M.; Vergez, L. M.; Land, M. L.; Motin, V. L.; Brubaker, R. R.; Fowler, J.; Hinnebusch, J.; Marceau, M.; Medigue, C.; Simonet, M.; Chenal-Francisque, V.; Souza, B.; Dacheux, D.; Elliott, J. M.; Derbise, A.; Hauser, L. J.; Garcia, E. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 13826−31. (18) Prentice, M. B.; Rahalison, L. Lancet 2007, 369, 1196−207. (19) Kenny, J. H.; Zhou, Y.; Schriefer, M. E.; Bearden, S. W. J. Microbiol. Methods 2008, 75, 293−301. (20) Matero, P.; Hemmila, H.; Tomaso, H.; Piiparinen, H.; Rantakokko-Jalava, K.; Nuotio, L.; Nikkari, S. Clin. Microbiol. Infect. 2011, 17, 34−43. (21) Simon, S.; Demeure, C.; Lamourette, P.; Filali, S.; Plaisance, M.; Creminon, C.; Volland, H.; Carniel, E. PLoS One 2013, 8, e54947. (22) Arbeli, Z.; Fuentes, C. L. FEMS Microbiol. Lett. 2007, 272, 269− 75. (23) Chenau, J.; Fenaille, F.; Ezan, E.; Morel, N.; Lamourette, P.; Goossens, P. L.; Becher, F. Anal. Chem. 2011, 83, 8675−82. (24) Tate, J.; Ward, G. Clin. Biochem. Rev. 2004, 25, 105−20. (25) Chanteau, S.; Rahalison, L.; Ralafiarisoa, L.; Foulon, J.; Ratsitorahina, M.; Ratsifasoamanana, L.; Carniel, E.; Nato, F. Lancet 2003, 361, 211−16. (26) Motin, V. L.; Georgescu, A. M.; Fitch, J. P.; Gu, P. P.; Nelson, D. O.; Mabery, S. L.; Garnham, J. B.; Sokhansanj, B. A.; Ott, L. L.; Coleman, M. A.; Elliott, J. M.; Kegelmeyer, L. M.; Wyrobek, A. J.; Slezak, T. R.; Brubaker, R. R.; Garcia, E. J. Bacteriol. 2004, 186, 6298− 305. (27) Du, Y.; Galyov, E.; Forsberg, A. Contrib. Microbiol. Immunol. 1995, 13, 321−24. (28) Demirev, P. A.; Fenselau, C. J. Mass Spectrom. 2008, 43, 1441− 57. (29) Lasch, P.; Drevinek, M.; Nattermann, H.; Grunow, R.; Stammler, M.; Dieckmann, R.; Schwecke, T.; Naumann, D. Anal. Chem. 2010, 82, 8464−75. (30) Ayyadurai, S.; Flaudrops, C.; Raoult, D.; Drancourt, M. BMC Microbiol. 2010, 10, 285. (31) Ber, R.; Mamroud, E.; Aftalion, M.; Tidhar, A.; Gur, D.; Flashner, Y.; Cohen, S. Appl. Environ. Microbiol. 2003, 69, 5787−92.

These data show that even in extremely complex samples with physical, chemical, and/or biological nuisances such as soil, the immuno-LC−SRM method is able unambiguously to detect low levels of Y. pestis. To our knowledge, this is the first demonstration of the direct detection of Y. pestis in soil samples.

■

CONCLUSION We have developed an immuno-LC−MS/MS assay for the sensitive and specific direct detection of Y. pestis in environmental and food samples, allowing an LOD as low as 2 × 104 CFU/mL in milk or tap water. The sensitivity appears similar to or even better than those of immunological assays (ELISA and dipstick),21 with the additional advantage of a high level of specificity conferred by the multiplex SRM detection. The immuno-LC−MS/MS assay also allowed Y. pestis detection in a complex environmental sample such as soil. Even if the LOD was slightly impacted, this constitutes the first demonstration that the bacteria can be directly detected in such samples, which is of considerable interest in evaluating Y. pestis persistence in the ground. Overall, our process is sensitive enough to get rid of any prior bacterial cultivation. The time to results might be further decreased using accelerated trypsin digestion protocols.50 The immuno-LC−MS/MS assay is intended to be implemented in regulatory laboratories for confirmation of results obtained with rapid field detection assays such as dipsticks. This method, which is primarily aimed at detecting Y. pestis in environmental samples, could also be applied to biological samples for plague diagnosis. The monitoring of F1 antigen, which is extensively produced during infection, could be easily implemented in the present approach.

■

ASSOCIATED CONTENT

* Supporting Information S

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org/.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: 33-1-69-08-13-15. Fax: 33-1-69-08-59-07. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This project was supported by the joint ministerial R&D program in CBRNE (chemical, biological, radiological, nuclear and high-yield explosives) risks. It was also partly funded by the French Institute of Health Surveillance (InVS).

■

REFERENCES

(1) Perry, R. D.; Fetherston, J. D. Clin. Microbiol. Rev. 1997, 10, 35− 66. (2) Butler, T. Clin. Infect. Dis. 2009, 49, 736−42. (3) Carniel, E. Med. Hist. Suppl. 2008, 115−22. (4) Inglesby, T. V.; Dennis, D. T.; Henderson, D. A.; Bartlett, J. G.; Ascher, M. S.; Eitzen, E.; Fine, A. D.; Friedlander, A. M.; Hauer, J.; Koerner, J. F.; Layton, M.; McDade, J.; Osterholm, M. T.; O’Toole, T.; Parker, G.; Perl, T. M.; Russell, P. K.; Schoch-Spana, M.; Tonat, K. JAMA, J. Am. Med. Assoc. 2000, 283, 2281−90. (5) Pohanka, M.; Skladal, P. Folia Microbiol. (Prague, Czech Repub.) 2009, 54, 263−72. (6) Riedel, S. Proc. (Baylor Univ. Med. Cent.) 2005, 18, 116−24. 6151

dx.doi.org/10.1021/ac501371r | Anal. Chem. 2014, 86, 6144−6152

Analytical Chemistry

Article

(32) Rollins, S. E.; Rollins, S. M.; Ryan, E. T. Am. J. Clin. Pathol. 2003, 119 Suppl, S78−S85. (33) Picotti, P.; Rinner, O.; Stallmach, R.; Dautel, F.; Farrah, T.; Domon, B.; Wenschuh, H.; Aebersold, R. Nat. Methods 2010, 7, 43− 46. (34) Madonna, A. J.; Basile, F.; Furlong, E.; Voorhees, K. J. Rapid Commun. Mass Spectrom. 2001, 15, 1068−74. (35) McGrath, S. C.; Schieltz, D. M.; McWilliams, L. G.; Pirkle, J. L.; Barr, J. R. Anal. Chem. 2011, 83, 2897−905. (36) Lasch, P.; Nattermann, H.; Erhard, M.; Stammler, M.; Grunow, R.; Bannert, N.; Appel, B.; Naumann, D. Anal. Chem. 2008, 80, 2026− 34. (37) Hixson, K. K.; Adkins, J. N.; Baker, S. E.; Moore, R. J.; Chromy, B. A.; Smith, R. D.; Cutchen-Maloney, S. L.; Lipton, M. S. J. Proteome. Res. 2006, 5, 3008−17. (38) Chenau, J.; Fenaille, F.; Caro, V.; Haustant, M.; Diancourt, L.; Klee, S. R.; Junot, C.; Ezan, E.; Goossens, P. L.; Becher, F. Mol. Cell Proteomics 2014, 13, 716−32. (39) Wunschel, D.; Engelmann, H.; Victry, K.; Clowers, B.; Sorensen, C.; Valentine, N.; Mahoney, C.; Wietsma, T.; Wahl, K. Mol. Cell Probes 2014, 28, 65−72. (40) Larson, M. A.; Ding, S. J.; Slater, S. R.; Hanway, A.; Bartling, A. M.; Fey, P. D.; Lockridge, O.; Francesconi, S. C.; Hinrichs, S. H. Proteomics Clin. Appl. 2013, 7, 416−23. (41) Hinnebusch, B. J.; Rudolph, A. E.; Cherepanov, P.; Dixon, J. E.; Schwan, T. G.; Forsberg, A. Science 2002, 296, 733−35. (42) Patzer, S. I.; Albrecht, R.; Braun, V.; Zeth, K. J. Biol. Chem. 2012, 287, 23381−96. (43) Sodeinde, O. A.; Subrahmanyam, Y. V.; Stark, K.; Quan, T.; Bao, Y.; Goguen, J. D. Science 1992, 258, 1004−07. (44) Suomalainen, M.; Haiko, J.; Ramu, P.; Lobo, L.; Kukkonen, M.; Westerlund-Wikstrom, B.; Virkola, R.; Lahteenmaki, K.; Korhonen, T. K. Adv. Exp. Med. Biol. 2007, 603, 268−78. (45) Gallien, S.; Duriez, E.; Domon, B. J. Mass Spectrom. 2011, 46, 298−312. (46) Haiko, J.; Laakkonen, L.; Juuti, K.; Kalkkinen, N.; Korhonen, T. K. J. Bacteriol. 2010, 192, 4553−61. (47) Haiko, J.; Laakkonen, L.; Westerlund-Wikstrom, B.; Korhonen, T. K. BMC Evol. Biol. 2011, 11, 43. (48) Felek, S.; Muszynski, A.; Carlson, R. W.; Tsang, T. M.; Hinnebusch, B. J.; Krukonis, E. S. Infect. Immun. 2010, 78, 1163−75. (49) Eisen, R. J.; Gage, K. L. Vet. Res. 2009, 40, 1. (50) Seyer, A.; Fenaille, F.; Feraudet-Tarisse, C.; Volland, H.; Popoff, M. R.; Tabet, J. C.; Junot, C.; Becher, F. Anal. Chem. 2012, 84, 5103− 09.

6152

dx.doi.org/10.1021/ac501371r | Anal. Chem. 2014, 86, 6144−6152