Anal Bioanal Chem DOI 10.1007/s00216-014-7710-2
RESEARCH PAPER
Identification and quantification of ricin in biomedical samples by magnetic immunocapture enrichment and liquid chromatography electrospray ionization tandem mass spectrometry Xiaoxi Ma & Jijun Tang & Chunzheng Li & Qin Liu & Jia Chen & Hua Li & Lei Guo & Jianwei Xie
Received: 31 October 2013 / Revised: 10 February 2014 / Accepted: 19 February 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Ricin is a toxic protein derived from castor beans and composed of a cytotoxic A chain and a galactose-binding B chain linked by a disulfide bond, which can inhibit protein synthesis and cause cell death. Owing to its high toxicity, ease of preparation, and lack of medical countermeasures, ricin has been listed as both chemical and biological warfare agents. For homeland security or public safety, the unambiguous, sensitive, and rapid methods for identification and quantification of ricin in complicated matrices are of urgent need. Mass spectrometric analysis, which provides specific and sensitive characterization of protein, can be applied to confirm and quantify ricin. Here, we report a liquid chromatographyelectrospray ionization tandem mass spectrometry (LC-ESIMS/MS) method in which ricin was extracted and enriched from serum by immunocapture using anti-ricin monoclonal antibody 3D74 linked to magnetic beads, then digested by trypsin, and analyzed by LC-ESI-MS/MS. Among 19 distinct peptides observed in LC-quadrupole/time of flight-MS (LCQTOF-MS), two specific and sensitive peptides, T 7A (49VGLPINQR56) and T14B (188DNCLTSDSNIR198), were chosen, and a highly sensitive determination of ricin was established in LC-triple quadrupole-MS (LC-QqQ-MS)
operating in multiple reaction monitoring mode. These specific peptides can definitely distinguish ricin from the homologous protein Ricinus communis agglutinin (RCA120), even though the amino acid sequence homology of the A-chain of ricin and RCA120 is up to ca. 93 % and that of B-chain is ca. 85 %. Furthermore, peptide T7A was preferred in the quantification of ricin because its sensitivity was at least one order of magnitude higher than that of the peptide T 14B. Combined with immunocapture enrichment, this method provided a limit of detection of ca. 2.5 ng/mL and the limit of quantification was ca. 5 ng/mL of ricin in serum, respectively. Both precision and accuracy of this method were determined and the RSD was less than 15 %. This established method was then applied to measure ricin in serum samples collected from rats exposed to ricin at the dosage of 50 μg/kg in an intravenous injection manner. The results showed that ca. 10 ng/mL of the residual ricin in poisoned rats serum could be detected even at 12 h after exposure.
Published in the topical collection Analysis of Chemicals Relevant to the Chemical Weapons Convention with guest editors Marc-Michael Blum and R. V. S. Murty Mamidanna.
Introduction
Xiaoxi Ma and Jijun Tang contributed equally to this work. X. Ma : J. Tang : C. Li : Q. Liu : J. Chen : H. Li : L. Guo (*) : J. Xie (*) State Key Laboratory of Toxicology and Medical Countermeasures, and Laboratory of Toxicant Analysis, Institute of Pharmacology and Toxicology, Academy of Military Medical Sciences, 27 Taiping Road, Haidian District, Beijing 100850, China e-mail: [email protected] e-mail: [email protected]
Keywords Ricin . Mass spectrometry . Immunocapture . Peptide . Biomedical sample
Ricin toxin, derived from the castor beans (Ricinus communis), is a class II ribosome inactivating protein (RIPII), which is extremely toxic to eukaryotic cells [1, 2]. It is composed of an enzymatically active A chain (ricin-A chain), a galactose residue-binding B chain (ricin-B chain), and a disulfide bond linkage. Ricin-B chain binds to eukaryotic cells by interacting with cell-surface galactosides, and carries the toxic ricin-A chain into the cells via endocytosis. After entering a cell, ricin-A chain exerts its ribosome-inactivating
X. Ma et al.
toxicity and inhibits protein synthesis, thus causing cell death. This cytotoxin has been tested as an immunotoxin in cancer chemotherapy; ricin has thereby attracted great caution as a potential threat of biological warfare agent because of its wide availability, ease of preparation, lack of valid antidotes, and high lethality. Ricin has been confirmed as the main cause in several assassination and terror events (e.g., the notorious “umbrella murder” event causing the death of Georgi Markov in 1978 [3], and the “ricin letter” terror plot that recently occurred targeting the White House in the USA, etc. For these reasons, ricin remains in place as the only protein in Schedule 1 of the Chemical Weapons Convention, and also listed in the categories of the Biological Weapons Convention. For public safety and biodefense purposes, it is important to establish an unambiguous, rapid, and sensitive identification and quantification method for ricin in complex matrixes. So far the most common and sensitive determination methods for ricin are immuno-based assays, such as enzyme-linked immunosorbent assay (ELISA) [4], radioimmunoassay [5], antibody-based biosensors [6, 7], and so on. Immuno-based assays are not always unambiguous in cases where nonspecific binding of antibodies towards other proteins with similar binding motifs lead to false positive. For example, common anti-ricin antibodies cannot discriminate ricin from protein R. communis agglutinin (RCA120, molecular weight of ca.120 kDa), which is also produced in castor beans, consists of two subunits, A and B chains in each subunit, and is highly homologous but much less toxic than ricin. The amino acid sequence homology of the A chain of ricin and RCA120 is up to 94 % and that of B chain is ca. 84 % [8]. Another RIP-II protein, abrin, derived from the seeds of the plant Abrus precatorius, has a similar structure and toxicological mechanism as ricin. Because of high sequence homology and structural similarity, current immunoassays are difficult to accurately distinguish these three proteins (i.e., ricin, RCA120, and abrin. There is urgent need for developing new methods to identify and qualify the presence of ricin toxin based on different principles. For accurate and sensitive determination of biomacromolecules, soft ionization-based mass spectrometric (MS) techniques, such as electrospray ionization (ESI)-MS or matrix-assisted laser desorption ionization (MALDI)-MS, have attracted more interest and have been increasingly employed. In recent years, several different MS methods (i.e., MALDI-TOF-MS [9] or LC-ESI-MS [10, 11]) have been applied for the determination of ricin, typically combined with a tryptic digestion approach. Although these methods achieve rapid, sensitive, and specific characterization or determination results, their main focus is on the analysis of simple samples, such as purified and intact toxin, crude seed extracts, and beverages. For complicated biomedical samples, such as serum or plasma, only one related MALDI-TOF-MS method coupled with immunocapture preconcentration shows a
possibility of detection of ricin at ng/mL level [9]. The difficulty in LC-ESI-MS determination of ricin in serum samples lies in the relatively poor sensitivity than those of immunobiological assays. The main cause is that complicated serum matrix, especial high-abundant serum proteins, will severely affect the detection if not effectively pretreated prior to LC-MS analysis. To address this issue, we adopted a sample pretreatment described by McGrath et al. [10] (i.e., monoclonal antibody-based purification and enrichment procedures) and established a robust, specific, and sensitive LC-ESI-MS/ MS quantification method for measuring ricin in biomedical samples. The immunocapture approach dramatically reduces the complexity of matrix, allowing the extracted target protein to be further digested and analyzed with high sensitivity (Scheme 1). This method was then successfully applied to the quantification of real poisoned serum samples collected from rats injected intravenously with ricin at different time points after intoxication. To the best of our knowledge, this is the first LC-MS/MS method with high sensitivity and specificity in providing precise quantification of ricin in vivo.
Experimental Safety considerations Ricin is an extremely toxic protein listed in Schedule 1 of the Chemical Weapons Convention, and all related experiments should strictly follow the operating safety guidelines. Materials and chemicals Streptavidin MagneSphere Paramagnetic Particles (SAmagnetic beads, MBs) were purchased from Promega (Madison, WI, USA) as a suspension of 1 mg of MBs/mL. Centricon centrifugal filter device with molecular weight cutoff of 30 kDa was purchased from Millipore (Billerica, MA, USA). The purified ricin was home-made according to the previous literatures [12] and its purity was greater than 95 % by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) analysis. The stock solution of ricin at 1 mg/mL was prepared in phosphate-buffered saline [(PBS), pH 7.4]. Anti-ricin monoclonal antibody (mAb) 3D74 was kindly donated by Professor Yuxia Wang at the Laboratory of Basic Medical Sciences, Beijing Institute of Pharmacology and Toxicology. The peptide KCFSRHLRPA, which comprised several unnatural amino acids (accurate molecular weight 1412.69) was synthesized in-house and used as internal standard (IS) during LC-ESI-MS/MS analysis. Trypsin (sequencing grade) was purchased from Promega (Madison, WI, USA). Biotin N-hydroxysuccinimide esters (NHS-Biotin), trifluoroacetic acid (TFA), formic acid (FA), dithiothreitol (DTT), and iodoacetamide (IAA) were obtained from
Identification and quantification of ricin
Scheme 1 Scheme of magnetic immunocapture assisted LC-MS/MS analysis of ricin
Sigma-Aldrich Co. (St. Louis, MO, USA). All other reagents were of analytical purity and purchased from Beijing Sinopharm Chemical Reagents Co. Ltd. (Beijing, China) unless specifically stated otherwise. Ultrapure water was produced in a Mill-Q water purification system (Millipore, Billerica, MA, USA). Animals and treatment Sprague-Dawley (SD) male rats weighing about 210 g were supplied by the Laboratory Animal Center of Beijing. All procedures towards animals were in compliance with the guidelines of the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC). The animals were housed in a controlled environment. Three SD rats were i.v.-injected with ricin diluted in 0.1 mL of saline solution at a dosage of 50 μg/kg. At 10, 20, 30 min, 1, 2, 4, 8, 12, and 16 h after ricin administration, the rats were anesthetized with diethyl ether, and the blood samples were drawn from orbit (eye socket). The serum was separated by centrifugation at 4 °C after blood coagulation and stored at –70 °C until analysis. Preparation of anti-ricin mAb coated magnetic beads The mAb was first biotinylated, an aliquot of 80 μL NHSbiotin solution (0.15 mg/mL in 20 % dimethylformamide) was added to 20 μL MAb 3D74 (3 mg/mL in PBS), and incubated
at room temperature for 4 h, then 5 μL of 0.1 mol/L NH4Cl solution was added to the mixture to quench the active NHSbiotin. The mixture was ultra-filtered three times using the Centricon centrifugal filter device (Millipore, Billerica, MA, USA) at 6000 rpm for 5 min, and the resulting filtrate was discarded. The biotinylated mAb 3D74 was then recovered and diluted with 100 μL PBS at the final concentration of ca. 0.6 mg/mL. An amount of 600 μg SA-MBs was transferred to a 1.5-mL Eppendorf tube and washed twice with 800 μL PBS. After removal of the supernatant by a magnetic particle concentrator (Promega, Madison, WI, USA), an aliquot of 40 μL biotinylated mAb was mixed with SA-MBs for 4 h at room temperature under gentle shaking. The supernatant was then discarded, and the produced mAb-coated MBs (mAb-MBs) were washed twice with 800 μL of PBS and resuspended in 400 μL of PBS containing 0.1 mg/mL BSA. Immunocapture enrichment and trypsin digestion All calibration standards examined here were obtained by serially diluting purified ricin in blank serum. Aliquots of 500 μL calibration standards or serum samples collected from poisoned SD rats were mixed with 20 μL mAb-MBs and incubated for 1 h at room temperature with gentle shaking. After removing supernatant, the mAb-MBs were then recovered and washed twice with 500 μL PBS to remove nonspecifically bound proteins, and then twice with 500 μL of ultrapure water to remove the remaining buffer. Ricin elution
X. Ma et al.
was accomplished by incubating the mAb-MBs with 50 μL of 0.1 % TFA for 10 min at room temperature. The supernatant was transferred into a new Eppendorf tube and evaporated to dryness by centrifugal evaporation (Labconco Corp., Kansas City, Missouri, USA). Ricin was then resuspended in 20 μL of 50 mmol/L ammonium bicarbonate. For reduction of the disulfide bonds, aliquot of 1 μL of 200 mmol/L DTT was added and incubated for 1 h at 37 °C. After that, alkylation of the thiol groups was achieved by adding 3 μL of 200 mmol/L IAA and incubated for 30 min in the dark. Subsequently, aliquot of 1 μL of 1 mg/mL trypsin solution was added and incubated for 3 h at 37 °C; then the same amount of trypsin was added again and incubated for another 3 h. The digestion was terminated by adding 1 μL of 10 % FA. Four μL of 1 μg/mL IS peptide was added to the samples, the mixture was then centrifuged for 3 min at 14,000 rpm, and the liquid supernatant was transferred to an autosampler vial for LC-MS/MS analysis. LC-QTOF-MS analysis To select the specific and sensitive peptides for determination of ricin, tryptic digests of purified ricin without IS were first analyzed by LC-QTOF-MS. Chromatographic separations were performed on a wide-pore (300 Å) Zorbax 300SB-C18 column (2.1 mm×100 mm, 3.5 μm) mounted on Agilent 1200 HPLC system. Mobile phases consisted of water/acetonitrile/ formic acid (95/5/0.1, v/v/v, as mobile phase A), and water/ acetonitrile/formic acid (5/95/0.1, v/v/v, as mobile phase B). Gradient profile was 0 % B held at 0 to 2 min, linearly increased to 60 % B over 38 min, and then increased to 100 % B over 10 min, and held another 10 min, for a total run time of 60 min. The flow rate was 0.2 mL/min, and the injection volume was 5 μL. The LC eluent was directly connected to an Agilent 6520 Accurate Mass QTOF-MS system. Instrument parameters include: capillary temperature, 350 °C; gas flow, 8 L/min; ESI spray voltage, 3.8 kV; fragmentor, 210 V; collision energy, 20 V. Nitrogen was used as nebulizer and collision gas. The scan range was m/z 300– 3000 for peptide mass mapping and m/z 100–2000 for product ion spectrum acquisition.
identity. Filtering criteria were protein score greater than 11 and peptide score greater than 7, respectively. LC-QqQ-MS/MS analysis Analyses were carried out on an Agilent 1200 Series Rapid Resolution LC (RRLC) system coupled with an Agilent 6430 Triple Quadrupole LC/MS with ESI mode. LC separations were performed on a wide-pore Zorbax 300SB-C18 column (1.0 mm×50 mm, 3.5 μm) with a Zorbax StableBond guard column (1.0 mm×17 mm, 5 μm) kept at 40 °C. Mobile phases are the same as that used in LC-QTOF-MS analysis, and gradient profile was, 5 % B held at 0 to 2 min, linearly increased to 30 % B over 2.5 min, and then increased to 40 % B over 1 min, to 90 % B over 0.5 min, and held another 2 min, for a total run time of 8 min. The flow rate was 0.15 mL/min, and the injection volume was 2 μL. MS detection was performed in a positive ESI mode at 3.5 kV. The nebulizer was 35 psi, gas temperature was 350 °C, and gas flow was 8 L/min. Quantification was performed in MRM (multiple selected reaction monitoring) mode.
Results and discussion Ricin has been deemed to be a potential chemical and biological warfare agent because of its characteristics. To meet the counterterrorism need, many MS-based methods with high specificity, accuracy, and sensitivity have been developed for the identification and determination of ricin in recent years, but most of them were more suitable for environmental or contaminated food samples [9–11]. The accurate quantification method for residual ricin in biomedical matrix is still a challenge. The immunocapture as a sample pretreatment technique could effectively extract the target molecule from complicated matrix, which has been demonstrated in the previous literature [9]. Thus, this present work was to establish a combined approach [i.e., magnetic immunocapture enrichment of ricin from serum matrix along with the determination by LC-QqQ-MS (MRM)], which allows specific and sensitive determination of ricin in biomedical samples. Characterization of mAb-MBs
Database search The raw QTOF-MS data were analyzed in an Agilent Spectrum Mill MS proteomics workbench using the Swiss-Prot protein database search engine. The peak lists were first created with Spectrum Mill Data Extractor program. The processed data were then subjected to database search against the SwissProt protein database. The search parameters include, database: Swiss-Prot; species: all; digest: trypsin; modification: carbamethylation; masses: monoisotopic; search mode:
The traditional coupling approach for antibodies and MBs is a covalent conjugation one, occurring between the primary amino groups of antibody and the carboxylic acid or tosyl groups on the surface of MBs [13, 14]. In this way, multiple reaction steps are involved, the intermediate state of reactive group is very labile, and the side-reaction (e.g., hydrolyzation) sometimes occupies the reacting course and deteriorates the conjugation efficiency. Additionally, the antibody coupled MBs need blocking of protein, such as BSA, to reduce the
Identification and quantification of ricin
Fig. 1 LC-QTOF chromatogram of tryptic digests of purified ricin
nonspecific adsorption and retain the activity of immobilized antibody. In an alternative approach, antibody can be coated to MBs with covalently bound protein G by virtue of the interaction between the antibody and protein G [9, 15]. In this way, the coating efficiency is greatly influenced by the affinity between these two molecules, which is not tight enough to be inclined to dissociate, especially in complicated matrix sample like serum. In this paper, we described an efficient and convenient way, anti-ricin mAb 3D74 was biotinylated first, and then attached to SA-MBs via a much stronger streptavidin-biotin interaction [16]. The determination results of biotinylated mAb beforeand-after binding with SA-MBs by bicinchoninic acid assay (BCA assay) indicated 1 mg of SA-MBs can bind ca. 55 μg of mAb. Then, mAb-MBs were incubated with different concentrations of ricin; the residual amounts of ricin were determined by ELISA. This characterization result showed that the binding capacity of mAb-MBs was ca. 16.5 μg of ricin per 1 mg MBs. According to this good binding capacity, we applied 20 μL of 1 mg/mL mAb-MBs to extract and enrich hundreds of ng of ricin from serum samples each time in the subsequent experiments. Moreover, since a previous immunoassay had demonstrated that mAb 3D74 can recognize the whole molecule of ricin [17], only intact toxin in serum could be captured and enriched by these mAb-MBs, which promises the integrity of target protein; then the toxin can be treated in the following digestion and LC-MS/MS analysis.
sequence. The tryptic digests of purified ricin were first analyzed by LC-QTOF-MS. Nineteen distinct peptides were observed with 43 % coverage of the amino acid sequences through Spectrum Mill search (Fig. 1). Potential tryptic digested peptides were identified from the amino acid sequence of ricin in the Swiss Prot database. As is wellknown, the amino acid sequence homology between ricin and RCA120 is up to ca. 93 % and ca. 84 % for A and B chains, respectively [18], and no reported antibodies can completely distinguish them. Thus, much attention should be taken in the selection of peptide sequences to avoid false positive detection results. Among all observed 19 peptides, T5A, T7A, T9A, T10A, T11A, T13A, T23A, T14B, T15B, T18B, T19B, and T20B (subscripts A and B refers A and B chain, respectively) were the specific peptides differentiating ricin from RCA120, and in which T7A, T11A, and T14B were initially
Selection of specific peptides for ricin detection MS-based analysis for tryptic digests of target proteins can easily distinguish individual protein on the basis of amino acid
Fig. 2 LC-QqQ-MS/MS chromatogram of marker peptides T7A, T14B, and IS peptide
X. Ma et al. Table 1 Parameters of transition pairs Peptides
Precursor ion (m/z)
Product ion (m/z)
Fragmentor (V)
CE (V)
T7A T14B IS
[448.8]2+ [647.8]2+ [707.6]2+
[627.3]+ [792.4]+ [829.4]+
110 140 180
15 20 35
chosen as specific peptides for the determination of ricin based on the following criteria: high enough response in MS, specific for ricin, and not containing amino acid residues susceptible to chemical modification, such as cysteine, methionine, etc. Further experiments showed that T7A and T11A had almost the same retention times on the LC column even under various chromatographic separation conditions, and the detection sensitivity of T7A was over four times higher than that of T11A. Eventually, two peptides, T7A and T14B individually originated from A and B chains of ricin, were selected as the specific peptides for the determination of ricin in serum samples. Based on the results from Basic Local Alignment Search Tool (BLAST) analysis of the non-redundant protein database of the National Center of Bioinformatics (NCBI), these peptides
Fig. 3 Product ion spectra of tryptic peptides T7A and T14B specific for the identification of ricin
allowed ricin to be easily distinguished from RCA120. Additionally, based on the two peptides, ricin could also be distinguished from abrin with similar structures and properties to ricin, which shared more than 50 % of the amino acid sequence homology with ricin. Meanwhile, the combined use of T7A and T14B would confirm the presence of sole, intact ricin. In this work, a self-synthesized peptide KCFSRHLRPA as the internal standard was adopted for the quantification of ricin in serum matrix in the case of good peak shape, suitable retention time, which could not interfere with the determination of ricin, and good MS response (Fig. 2). Since this peptide consists of three trypsin cleavage sites (i.e., K, R, and R), the IS solution was added to the samples after termination of tryptic digestion. Then, the peptide markers T7A and T14B as well as IS were further optimized for detection parameters by LC-QqQ-MS operating in precursor ion scan and product ion scan modes, respectively. The optimized detection parameters for the three peptides are summarized in Table 1, and the product ions spectra of the double charged T7A and T14B are shown in Fig. 3. For the peptide T7A, the y-ion series product ions including y1, y3, y4, y5, y6, and y7 were observed together with b2. The T14B peptide generated the y1, y4, y5, y6, y7, y8, b2 and b4 product ions. According to the mass spectrometric
Identification and quantification of ricin
responses, the precursor/product ion pairs of m/z 448.82+→m/ z 627.3+, m/z 627.82+→m/z 792.4+, and m/z 707.62+→m/z 829.4+ were used as MRM transitions of T7A, T14B, and IS, respectively. Optimization of magnetic immunocapture enrichment and tryptic digestion Complicated biomedical matrix such as serum raises a great challenge in determining extremely low concentrations of target proteins; many high-abundant proteins result in a relatively poor recovery of ricin from serum. Actually, the detection of low abundant proteins in LC-MS analysis is hardly achieved without sample pretreatment. To date, one of the most efficient pretreatment ways for overcoming this challenge is immuno-based approach, especially when coupled
Fig. 5 The calibration curve of different concentrations of ricin spiked in blank serum (a), and the trend of residual ricin in serum of poisoned rats over time (b)
with magnetic separation by virtue of a much easier manipulation. In this paper, mAb 3D74 coupled with MBs were employed to capture ricin from serum matrices, and the immuno-captured ricin was then eluted by buffers including 5 % formic acid (v/v) or 0.1 % TFA (v/v). The results from MS quantification showed that the area ratios of specific peptides versus IS were highest under elution with 0.1 % TFA. The eluate was evaporated to dryness to eliminate the destructive effects of TFA on the subsequently tryptic digestion and LCMS/MS analysis. Results of tryptic digestion under different conditions indicated that trypsin solution should be added twice and the mixture should be incubated for 6 h. Extension
Table 2 Accuracy and precision of ricin in serum matrix in LC-QqQMS/MS, with a quantification peptide T7A (n=3) Added (ng/mL)
Found (ng/mL)
Accuracya (%)
Precisionb (%)
10 50 80
7.8±0.38 36.5±5.2 70±3.4
78 73 87
4.8 14.2 5
a
Fig. 4 LC-QqQ-MS/MS chromatograms of blank serum (a) and ricin spiked at 2.5 ng/mL (b) and 5 ng/mL, (c) into blank serum, respectively
Expressed as [(mean found concentrations/added concentrations) × 100].
b
Expressed as the relative standard deviation (RSD).
X. Ma et al.
of incubation time to 16 h did not remarkably gain higher peptide yields than at 6 h. Method performance The method performance was evaluated by assessing the quantification results of ricin spiked in blank serum. For generating a calibration curve, purified ricin was serially diluted into blank serum to make samples with different concentrations, then these samples were incubated with mAb-MBs, and the captured ricin was eluted from MBs, followed by trypsin digestion and LC-QqQ-MS analysis. Peptide T7A was found to offer the best sensitivity and thus employed as the quantification peptide, with a limit of detection (LOD) of 2.5 ng/mL (S/N is greater than 3), and the limit of quantification (LOQ) was 5 ng/mL (S/N is greater than 10) (Fig. 4). Although this sensitivity was only comparable to the LOD of conventional ELISA, our method established herein could provide an unambiguous identification result at the same time. The calibration curve had a linear response from 5 to 80 ng/mL with a linear equation y=0.005x–0.014 (r2 =0.99, Fig. 5a). Serum samples containing 10, 20, and 50 ng/mL of ricin were analyzed in triplicate, and the accuracy values at low, middle, and high levels were 78 %, 73 %, and 87 %, respectively, with precision values ranging from 4.8 % to 14.2 % (Table 2). It was believed that the accuracy of the measurement could be further improved through isotope dilution mass spectrometry. Determination of residual ricin in serum samples of poisoned rats In case of terrorist attacks of ricin or accidental ingestion of castor beans, it is of great importance to analyze biomedical samples. To test the robustness of our method, the contents of residual ricin in rat serum were measured, which were collected at different times after i.v. administration of ricin to rats at a dosage level of 50 μg/kg (ca. 10 LD50). The trend of residual ricin over time was plotted as using peptide T7A for quantification (Fig. 5b). The results showed that both T7A and T14B (data not shown) appeared within 2 h after intoxication, and T7A could be detected up to 12 h. After that, the amount of residual ricin was lower than the LOD of this method. The determined amount of ricin in serum was much lower than that of administration because of the following reasons: a substantial part of ricin was delivered into various tissues via blood flow, and the antibody mAb 3D74 could only bind intact ricin, and some divided ricin A- or B-chains might be missed.
Conclusions The sensitive and quantitative LC-ESI-MS/MS methods were developed for identification and quantification of ricin in
serum samples. In this method, a magnetic immunocapture pretreatment based on specific mAb-coated MBs was implemented in order to extract and enrich target biotoxin from complicated matrices. Although comparing with current immuno-based bioassays, the LOD of this method was not quite superior, the accurate MS analysis of this work provided a great advantage of unambiguous identification because of the simultaneous monitoring of characteristic peptides of ricin. Therefore, ricin can be accurately distinguished from RCA120 through specific peptides T7A and T14B. Finally, the method was successfully employed to determine residual ricin in serum of poisoned rats. The residual ricin in serum of poisoned rats can be detected until 12 h later at a concentration of ca. 10 ng/mL. Acknowledgments This research was supported by grants from the National Natural Science Foundation of China (grant nos. 21175152, 81001262, and 81202248), and National Science and Technology Major Project of the Ministry of Science and Technology of China (grant no. 2012ZX09301003-001-010).
References 1. Barbieri L, Battelli MG, Stirpe F (1993) Ribosome-inactivating proteins from plants. Biochim Biophys Acta 1154:237–282 2. Lord JM, Roberts LM, Robertus JD (1994) Ricin: structure, mode of action, and some current applications. FASEB J 8:201–208 3. Papaloucas M, Papaloucas C, Stergioulas A (2008) Ricin and assassination of Georgi Markov Pakistan. J Biol Sci 11:2370–2371 4. Ma X, Liu H, Tang J, Guo L, Xie J (2011) Comparison of three enzyme-linked immunosorbent assay methods for quantitative determination of ricin. Chin J Anal Chem 39:685–689 5. Godal A, Olsnes S, Pihl A (1981) Radioimmunoassays of abrin and ricin in blood. J Toxicol Environ Health 8:409–417 6. Ligler FS, Taitt CR, Shriver-Lake LC, Sapsford KE, Shubin Y, Golden JP (2003) Array biosensor for detection of toxins. Anal Bioanal Chem 377:469–477 7. Goldman ER, Clapp AR, Anderson GP, Uyeda HT, Mauro JM, Medintz IL, Mattoussi H (2004) Multiplexed toxin analysis using four colors of quantum dot fluorore agents. Anal Chem 76:684–688 8. Araki F, Funatsu G (1987) The complete amino acid sequence of the B-chain of ricin E isolated from small-grain castor bean seeds. Biochim Biophys Acta 911:191–200 9. Duriez E, Fenaille F, Tabet JC, Lamourette P, Hilaire D, Becher F, Ezan E (2008) Detection of ricin in complex samples by immunocapture and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. J Proteome Res 7:4154–4163 10. McGrath SC, Schieltz DM, McWilliams LG, Pirkle JL, Barr JR (2011) Detection and quantification of ricin in beverages using isotope dilution tandem mass spectrometry. Anal Chem 83:2897–2905 11. Ostin A, Bergstrom T, Fredriksson SA, Nilsson C (2007) Solventassisted trypsin digestion of ricin for forensic identification by LCESI MS/MS. Anal Chem 79:6271–6278 12. Tang J, Xie J, Shao N, Yan Y (2006) The DNA aptamers that specifically recognize ricin toxin are selected by two in vitro selection methods. Electrophoresis 27:1303–1311 13. Skibeli V, Nissen-Lie G, Torjesen P (2001) Sugar profiling proves that human serum erythropoietin differs from recombinant human erythropoietin. Blood 98:3626–3634
Identification and quantification of ricin 14. Guan F, Uboh CE, Soma LR, Birks E, Chen J, Mitchell J, You Y, Rudy J, Xu F, Li X, Mbuy G (2007) LC-MS/MS method for confirmation of recombinant human erythropoietin and darbepoetin α in equine plasma. Anal Chem 79:4627–4635 15. Becher F, Duriez E, Volland H, Tabet JC, Ezan E (2007) Detection of functional ricin by immunoaffinity and liquid chromatography tandem mass spectrometry. Anal Chem 79:659–665 16. Guo L, Wei Y, Xie J (2011) Analysis of ricin: immunocapture. In: Vanninen P (ed) Recommended operating procedures for analysis in
the verification of chemical disarmament. University of Helsinki, Helsinki, pp 419–423 17. Men J, Lang L, Wang C, Wu J, Zhao Y, Jia P, Wei W, Wang Y (2010) Detection of residual toxin in tissues of ricin-poisoned mice by sandwich enzyme-linked immunosorbent assay and immunoprecipitation. Anal Biochem 401:211–216 18. Araki F, Funatsu G (1987) The complete amino acid sequence of the B-chain of ricin E isolated from small-grain castor bean seeds. Biochim Biophys Acta 911:562–568
RESEARCH PAPER
Identification and quantification of ricin in biomedical samples by magnetic immunocapture enrichment and liquid chromatography electrospray ionization tandem mass spectrometry Xiaoxi Ma & Jijun Tang & Chunzheng Li & Qin Liu & Jia Chen & Hua Li & Lei Guo & Jianwei Xie
Received: 31 October 2013 / Revised: 10 February 2014 / Accepted: 19 February 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Ricin is a toxic protein derived from castor beans and composed of a cytotoxic A chain and a galactose-binding B chain linked by a disulfide bond, which can inhibit protein synthesis and cause cell death. Owing to its high toxicity, ease of preparation, and lack of medical countermeasures, ricin has been listed as both chemical and biological warfare agents. For homeland security or public safety, the unambiguous, sensitive, and rapid methods for identification and quantification of ricin in complicated matrices are of urgent need. Mass spectrometric analysis, which provides specific and sensitive characterization of protein, can be applied to confirm and quantify ricin. Here, we report a liquid chromatographyelectrospray ionization tandem mass spectrometry (LC-ESIMS/MS) method in which ricin was extracted and enriched from serum by immunocapture using anti-ricin monoclonal antibody 3D74 linked to magnetic beads, then digested by trypsin, and analyzed by LC-ESI-MS/MS. Among 19 distinct peptides observed in LC-quadrupole/time of flight-MS (LCQTOF-MS), two specific and sensitive peptides, T 7A (49VGLPINQR56) and T14B (188DNCLTSDSNIR198), were chosen, and a highly sensitive determination of ricin was established in LC-triple quadrupole-MS (LC-QqQ-MS)
operating in multiple reaction monitoring mode. These specific peptides can definitely distinguish ricin from the homologous protein Ricinus communis agglutinin (RCA120), even though the amino acid sequence homology of the A-chain of ricin and RCA120 is up to ca. 93 % and that of B-chain is ca. 85 %. Furthermore, peptide T7A was preferred in the quantification of ricin because its sensitivity was at least one order of magnitude higher than that of the peptide T 14B. Combined with immunocapture enrichment, this method provided a limit of detection of ca. 2.5 ng/mL and the limit of quantification was ca. 5 ng/mL of ricin in serum, respectively. Both precision and accuracy of this method were determined and the RSD was less than 15 %. This established method was then applied to measure ricin in serum samples collected from rats exposed to ricin at the dosage of 50 μg/kg in an intravenous injection manner. The results showed that ca. 10 ng/mL of the residual ricin in poisoned rats serum could be detected even at 12 h after exposure.
Published in the topical collection Analysis of Chemicals Relevant to the Chemical Weapons Convention with guest editors Marc-Michael Blum and R. V. S. Murty Mamidanna.
Introduction
Xiaoxi Ma and Jijun Tang contributed equally to this work. X. Ma : J. Tang : C. Li : Q. Liu : J. Chen : H. Li : L. Guo (*) : J. Xie (*) State Key Laboratory of Toxicology and Medical Countermeasures, and Laboratory of Toxicant Analysis, Institute of Pharmacology and Toxicology, Academy of Military Medical Sciences, 27 Taiping Road, Haidian District, Beijing 100850, China e-mail: [email protected] e-mail: [email protected]
Keywords Ricin . Mass spectrometry . Immunocapture . Peptide . Biomedical sample
Ricin toxin, derived from the castor beans (Ricinus communis), is a class II ribosome inactivating protein (RIPII), which is extremely toxic to eukaryotic cells [1, 2]. It is composed of an enzymatically active A chain (ricin-A chain), a galactose residue-binding B chain (ricin-B chain), and a disulfide bond linkage. Ricin-B chain binds to eukaryotic cells by interacting with cell-surface galactosides, and carries the toxic ricin-A chain into the cells via endocytosis. After entering a cell, ricin-A chain exerts its ribosome-inactivating
X. Ma et al.
toxicity and inhibits protein synthesis, thus causing cell death. This cytotoxin has been tested as an immunotoxin in cancer chemotherapy; ricin has thereby attracted great caution as a potential threat of biological warfare agent because of its wide availability, ease of preparation, lack of valid antidotes, and high lethality. Ricin has been confirmed as the main cause in several assassination and terror events (e.g., the notorious “umbrella murder” event causing the death of Georgi Markov in 1978 [3], and the “ricin letter” terror plot that recently occurred targeting the White House in the USA, etc. For these reasons, ricin remains in place as the only protein in Schedule 1 of the Chemical Weapons Convention, and also listed in the categories of the Biological Weapons Convention. For public safety and biodefense purposes, it is important to establish an unambiguous, rapid, and sensitive identification and quantification method for ricin in complex matrixes. So far the most common and sensitive determination methods for ricin are immuno-based assays, such as enzyme-linked immunosorbent assay (ELISA) [4], radioimmunoassay [5], antibody-based biosensors [6, 7], and so on. Immuno-based assays are not always unambiguous in cases where nonspecific binding of antibodies towards other proteins with similar binding motifs lead to false positive. For example, common anti-ricin antibodies cannot discriminate ricin from protein R. communis agglutinin (RCA120, molecular weight of ca.120 kDa), which is also produced in castor beans, consists of two subunits, A and B chains in each subunit, and is highly homologous but much less toxic than ricin. The amino acid sequence homology of the A chain of ricin and RCA120 is up to 94 % and that of B chain is ca. 84 % [8]. Another RIP-II protein, abrin, derived from the seeds of the plant Abrus precatorius, has a similar structure and toxicological mechanism as ricin. Because of high sequence homology and structural similarity, current immunoassays are difficult to accurately distinguish these three proteins (i.e., ricin, RCA120, and abrin. There is urgent need for developing new methods to identify and qualify the presence of ricin toxin based on different principles. For accurate and sensitive determination of biomacromolecules, soft ionization-based mass spectrometric (MS) techniques, such as electrospray ionization (ESI)-MS or matrix-assisted laser desorption ionization (MALDI)-MS, have attracted more interest and have been increasingly employed. In recent years, several different MS methods (i.e., MALDI-TOF-MS [9] or LC-ESI-MS [10, 11]) have been applied for the determination of ricin, typically combined with a tryptic digestion approach. Although these methods achieve rapid, sensitive, and specific characterization or determination results, their main focus is on the analysis of simple samples, such as purified and intact toxin, crude seed extracts, and beverages. For complicated biomedical samples, such as serum or plasma, only one related MALDI-TOF-MS method coupled with immunocapture preconcentration shows a
possibility of detection of ricin at ng/mL level [9]. The difficulty in LC-ESI-MS determination of ricin in serum samples lies in the relatively poor sensitivity than those of immunobiological assays. The main cause is that complicated serum matrix, especial high-abundant serum proteins, will severely affect the detection if not effectively pretreated prior to LC-MS analysis. To address this issue, we adopted a sample pretreatment described by McGrath et al. [10] (i.e., monoclonal antibody-based purification and enrichment procedures) and established a robust, specific, and sensitive LC-ESI-MS/ MS quantification method for measuring ricin in biomedical samples. The immunocapture approach dramatically reduces the complexity of matrix, allowing the extracted target protein to be further digested and analyzed with high sensitivity (Scheme 1). This method was then successfully applied to the quantification of real poisoned serum samples collected from rats injected intravenously with ricin at different time points after intoxication. To the best of our knowledge, this is the first LC-MS/MS method with high sensitivity and specificity in providing precise quantification of ricin in vivo.
Experimental Safety considerations Ricin is an extremely toxic protein listed in Schedule 1 of the Chemical Weapons Convention, and all related experiments should strictly follow the operating safety guidelines. Materials and chemicals Streptavidin MagneSphere Paramagnetic Particles (SAmagnetic beads, MBs) were purchased from Promega (Madison, WI, USA) as a suspension of 1 mg of MBs/mL. Centricon centrifugal filter device with molecular weight cutoff of 30 kDa was purchased from Millipore (Billerica, MA, USA). The purified ricin was home-made according to the previous literatures [12] and its purity was greater than 95 % by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) analysis. The stock solution of ricin at 1 mg/mL was prepared in phosphate-buffered saline [(PBS), pH 7.4]. Anti-ricin monoclonal antibody (mAb) 3D74 was kindly donated by Professor Yuxia Wang at the Laboratory of Basic Medical Sciences, Beijing Institute of Pharmacology and Toxicology. The peptide KCFSRHLRPA, which comprised several unnatural amino acids (accurate molecular weight 1412.69) was synthesized in-house and used as internal standard (IS) during LC-ESI-MS/MS analysis. Trypsin (sequencing grade) was purchased from Promega (Madison, WI, USA). Biotin N-hydroxysuccinimide esters (NHS-Biotin), trifluoroacetic acid (TFA), formic acid (FA), dithiothreitol (DTT), and iodoacetamide (IAA) were obtained from
Identification and quantification of ricin
Scheme 1 Scheme of magnetic immunocapture assisted LC-MS/MS analysis of ricin
Sigma-Aldrich Co. (St. Louis, MO, USA). All other reagents were of analytical purity and purchased from Beijing Sinopharm Chemical Reagents Co. Ltd. (Beijing, China) unless specifically stated otherwise. Ultrapure water was produced in a Mill-Q water purification system (Millipore, Billerica, MA, USA). Animals and treatment Sprague-Dawley (SD) male rats weighing about 210 g were supplied by the Laboratory Animal Center of Beijing. All procedures towards animals were in compliance with the guidelines of the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC). The animals were housed in a controlled environment. Three SD rats were i.v.-injected with ricin diluted in 0.1 mL of saline solution at a dosage of 50 μg/kg. At 10, 20, 30 min, 1, 2, 4, 8, 12, and 16 h after ricin administration, the rats were anesthetized with diethyl ether, and the blood samples were drawn from orbit (eye socket). The serum was separated by centrifugation at 4 °C after blood coagulation and stored at –70 °C until analysis. Preparation of anti-ricin mAb coated magnetic beads The mAb was first biotinylated, an aliquot of 80 μL NHSbiotin solution (0.15 mg/mL in 20 % dimethylformamide) was added to 20 μL MAb 3D74 (3 mg/mL in PBS), and incubated
at room temperature for 4 h, then 5 μL of 0.1 mol/L NH4Cl solution was added to the mixture to quench the active NHSbiotin. The mixture was ultra-filtered three times using the Centricon centrifugal filter device (Millipore, Billerica, MA, USA) at 6000 rpm for 5 min, and the resulting filtrate was discarded. The biotinylated mAb 3D74 was then recovered and diluted with 100 μL PBS at the final concentration of ca. 0.6 mg/mL. An amount of 600 μg SA-MBs was transferred to a 1.5-mL Eppendorf tube and washed twice with 800 μL PBS. After removal of the supernatant by a magnetic particle concentrator (Promega, Madison, WI, USA), an aliquot of 40 μL biotinylated mAb was mixed with SA-MBs for 4 h at room temperature under gentle shaking. The supernatant was then discarded, and the produced mAb-coated MBs (mAb-MBs) were washed twice with 800 μL of PBS and resuspended in 400 μL of PBS containing 0.1 mg/mL BSA. Immunocapture enrichment and trypsin digestion All calibration standards examined here were obtained by serially diluting purified ricin in blank serum. Aliquots of 500 μL calibration standards or serum samples collected from poisoned SD rats were mixed with 20 μL mAb-MBs and incubated for 1 h at room temperature with gentle shaking. After removing supernatant, the mAb-MBs were then recovered and washed twice with 500 μL PBS to remove nonspecifically bound proteins, and then twice with 500 μL of ultrapure water to remove the remaining buffer. Ricin elution
X. Ma et al.
was accomplished by incubating the mAb-MBs with 50 μL of 0.1 % TFA for 10 min at room temperature. The supernatant was transferred into a new Eppendorf tube and evaporated to dryness by centrifugal evaporation (Labconco Corp., Kansas City, Missouri, USA). Ricin was then resuspended in 20 μL of 50 mmol/L ammonium bicarbonate. For reduction of the disulfide bonds, aliquot of 1 μL of 200 mmol/L DTT was added and incubated for 1 h at 37 °C. After that, alkylation of the thiol groups was achieved by adding 3 μL of 200 mmol/L IAA and incubated for 30 min in the dark. Subsequently, aliquot of 1 μL of 1 mg/mL trypsin solution was added and incubated for 3 h at 37 °C; then the same amount of trypsin was added again and incubated for another 3 h. The digestion was terminated by adding 1 μL of 10 % FA. Four μL of 1 μg/mL IS peptide was added to the samples, the mixture was then centrifuged for 3 min at 14,000 rpm, and the liquid supernatant was transferred to an autosampler vial for LC-MS/MS analysis. LC-QTOF-MS analysis To select the specific and sensitive peptides for determination of ricin, tryptic digests of purified ricin without IS were first analyzed by LC-QTOF-MS. Chromatographic separations were performed on a wide-pore (300 Å) Zorbax 300SB-C18 column (2.1 mm×100 mm, 3.5 μm) mounted on Agilent 1200 HPLC system. Mobile phases consisted of water/acetonitrile/ formic acid (95/5/0.1, v/v/v, as mobile phase A), and water/ acetonitrile/formic acid (5/95/0.1, v/v/v, as mobile phase B). Gradient profile was 0 % B held at 0 to 2 min, linearly increased to 60 % B over 38 min, and then increased to 100 % B over 10 min, and held another 10 min, for a total run time of 60 min. The flow rate was 0.2 mL/min, and the injection volume was 5 μL. The LC eluent was directly connected to an Agilent 6520 Accurate Mass QTOF-MS system. Instrument parameters include: capillary temperature, 350 °C; gas flow, 8 L/min; ESI spray voltage, 3.8 kV; fragmentor, 210 V; collision energy, 20 V. Nitrogen was used as nebulizer and collision gas. The scan range was m/z 300– 3000 for peptide mass mapping and m/z 100–2000 for product ion spectrum acquisition.
identity. Filtering criteria were protein score greater than 11 and peptide score greater than 7, respectively. LC-QqQ-MS/MS analysis Analyses were carried out on an Agilent 1200 Series Rapid Resolution LC (RRLC) system coupled with an Agilent 6430 Triple Quadrupole LC/MS with ESI mode. LC separations were performed on a wide-pore Zorbax 300SB-C18 column (1.0 mm×50 mm, 3.5 μm) with a Zorbax StableBond guard column (1.0 mm×17 mm, 5 μm) kept at 40 °C. Mobile phases are the same as that used in LC-QTOF-MS analysis, and gradient profile was, 5 % B held at 0 to 2 min, linearly increased to 30 % B over 2.5 min, and then increased to 40 % B over 1 min, to 90 % B over 0.5 min, and held another 2 min, for a total run time of 8 min. The flow rate was 0.15 mL/min, and the injection volume was 2 μL. MS detection was performed in a positive ESI mode at 3.5 kV. The nebulizer was 35 psi, gas temperature was 350 °C, and gas flow was 8 L/min. Quantification was performed in MRM (multiple selected reaction monitoring) mode.
Results and discussion Ricin has been deemed to be a potential chemical and biological warfare agent because of its characteristics. To meet the counterterrorism need, many MS-based methods with high specificity, accuracy, and sensitivity have been developed for the identification and determination of ricin in recent years, but most of them were more suitable for environmental or contaminated food samples [9–11]. The accurate quantification method for residual ricin in biomedical matrix is still a challenge. The immunocapture as a sample pretreatment technique could effectively extract the target molecule from complicated matrix, which has been demonstrated in the previous literature [9]. Thus, this present work was to establish a combined approach [i.e., magnetic immunocapture enrichment of ricin from serum matrix along with the determination by LC-QqQ-MS (MRM)], which allows specific and sensitive determination of ricin in biomedical samples. Characterization of mAb-MBs
Database search The raw QTOF-MS data were analyzed in an Agilent Spectrum Mill MS proteomics workbench using the Swiss-Prot protein database search engine. The peak lists were first created with Spectrum Mill Data Extractor program. The processed data were then subjected to database search against the SwissProt protein database. The search parameters include, database: Swiss-Prot; species: all; digest: trypsin; modification: carbamethylation; masses: monoisotopic; search mode:
The traditional coupling approach for antibodies and MBs is a covalent conjugation one, occurring between the primary amino groups of antibody and the carboxylic acid or tosyl groups on the surface of MBs [13, 14]. In this way, multiple reaction steps are involved, the intermediate state of reactive group is very labile, and the side-reaction (e.g., hydrolyzation) sometimes occupies the reacting course and deteriorates the conjugation efficiency. Additionally, the antibody coupled MBs need blocking of protein, such as BSA, to reduce the
Identification and quantification of ricin
Fig. 1 LC-QTOF chromatogram of tryptic digests of purified ricin
nonspecific adsorption and retain the activity of immobilized antibody. In an alternative approach, antibody can be coated to MBs with covalently bound protein G by virtue of the interaction between the antibody and protein G [9, 15]. In this way, the coating efficiency is greatly influenced by the affinity between these two molecules, which is not tight enough to be inclined to dissociate, especially in complicated matrix sample like serum. In this paper, we described an efficient and convenient way, anti-ricin mAb 3D74 was biotinylated first, and then attached to SA-MBs via a much stronger streptavidin-biotin interaction [16]. The determination results of biotinylated mAb beforeand-after binding with SA-MBs by bicinchoninic acid assay (BCA assay) indicated 1 mg of SA-MBs can bind ca. 55 μg of mAb. Then, mAb-MBs were incubated with different concentrations of ricin; the residual amounts of ricin were determined by ELISA. This characterization result showed that the binding capacity of mAb-MBs was ca. 16.5 μg of ricin per 1 mg MBs. According to this good binding capacity, we applied 20 μL of 1 mg/mL mAb-MBs to extract and enrich hundreds of ng of ricin from serum samples each time in the subsequent experiments. Moreover, since a previous immunoassay had demonstrated that mAb 3D74 can recognize the whole molecule of ricin [17], only intact toxin in serum could be captured and enriched by these mAb-MBs, which promises the integrity of target protein; then the toxin can be treated in the following digestion and LC-MS/MS analysis.
sequence. The tryptic digests of purified ricin were first analyzed by LC-QTOF-MS. Nineteen distinct peptides were observed with 43 % coverage of the amino acid sequences through Spectrum Mill search (Fig. 1). Potential tryptic digested peptides were identified from the amino acid sequence of ricin in the Swiss Prot database. As is wellknown, the amino acid sequence homology between ricin and RCA120 is up to ca. 93 % and ca. 84 % for A and B chains, respectively [18], and no reported antibodies can completely distinguish them. Thus, much attention should be taken in the selection of peptide sequences to avoid false positive detection results. Among all observed 19 peptides, T5A, T7A, T9A, T10A, T11A, T13A, T23A, T14B, T15B, T18B, T19B, and T20B (subscripts A and B refers A and B chain, respectively) were the specific peptides differentiating ricin from RCA120, and in which T7A, T11A, and T14B were initially
Selection of specific peptides for ricin detection MS-based analysis for tryptic digests of target proteins can easily distinguish individual protein on the basis of amino acid
Fig. 2 LC-QqQ-MS/MS chromatogram of marker peptides T7A, T14B, and IS peptide
X. Ma et al. Table 1 Parameters of transition pairs Peptides
Precursor ion (m/z)
Product ion (m/z)
Fragmentor (V)
CE (V)
T7A T14B IS
[448.8]2+ [647.8]2+ [707.6]2+
[627.3]+ [792.4]+ [829.4]+
110 140 180
15 20 35
chosen as specific peptides for the determination of ricin based on the following criteria: high enough response in MS, specific for ricin, and not containing amino acid residues susceptible to chemical modification, such as cysteine, methionine, etc. Further experiments showed that T7A and T11A had almost the same retention times on the LC column even under various chromatographic separation conditions, and the detection sensitivity of T7A was over four times higher than that of T11A. Eventually, two peptides, T7A and T14B individually originated from A and B chains of ricin, were selected as the specific peptides for the determination of ricin in serum samples. Based on the results from Basic Local Alignment Search Tool (BLAST) analysis of the non-redundant protein database of the National Center of Bioinformatics (NCBI), these peptides
Fig. 3 Product ion spectra of tryptic peptides T7A and T14B specific for the identification of ricin
allowed ricin to be easily distinguished from RCA120. Additionally, based on the two peptides, ricin could also be distinguished from abrin with similar structures and properties to ricin, which shared more than 50 % of the amino acid sequence homology with ricin. Meanwhile, the combined use of T7A and T14B would confirm the presence of sole, intact ricin. In this work, a self-synthesized peptide KCFSRHLRPA as the internal standard was adopted for the quantification of ricin in serum matrix in the case of good peak shape, suitable retention time, which could not interfere with the determination of ricin, and good MS response (Fig. 2). Since this peptide consists of three trypsin cleavage sites (i.e., K, R, and R), the IS solution was added to the samples after termination of tryptic digestion. Then, the peptide markers T7A and T14B as well as IS were further optimized for detection parameters by LC-QqQ-MS operating in precursor ion scan and product ion scan modes, respectively. The optimized detection parameters for the three peptides are summarized in Table 1, and the product ions spectra of the double charged T7A and T14B are shown in Fig. 3. For the peptide T7A, the y-ion series product ions including y1, y3, y4, y5, y6, and y7 were observed together with b2. The T14B peptide generated the y1, y4, y5, y6, y7, y8, b2 and b4 product ions. According to the mass spectrometric
Identification and quantification of ricin
responses, the precursor/product ion pairs of m/z 448.82+→m/ z 627.3+, m/z 627.82+→m/z 792.4+, and m/z 707.62+→m/z 829.4+ were used as MRM transitions of T7A, T14B, and IS, respectively. Optimization of magnetic immunocapture enrichment and tryptic digestion Complicated biomedical matrix such as serum raises a great challenge in determining extremely low concentrations of target proteins; many high-abundant proteins result in a relatively poor recovery of ricin from serum. Actually, the detection of low abundant proteins in LC-MS analysis is hardly achieved without sample pretreatment. To date, one of the most efficient pretreatment ways for overcoming this challenge is immuno-based approach, especially when coupled
Fig. 5 The calibration curve of different concentrations of ricin spiked in blank serum (a), and the trend of residual ricin in serum of poisoned rats over time (b)
with magnetic separation by virtue of a much easier manipulation. In this paper, mAb 3D74 coupled with MBs were employed to capture ricin from serum matrices, and the immuno-captured ricin was then eluted by buffers including 5 % formic acid (v/v) or 0.1 % TFA (v/v). The results from MS quantification showed that the area ratios of specific peptides versus IS were highest under elution with 0.1 % TFA. The eluate was evaporated to dryness to eliminate the destructive effects of TFA on the subsequently tryptic digestion and LCMS/MS analysis. Results of tryptic digestion under different conditions indicated that trypsin solution should be added twice and the mixture should be incubated for 6 h. Extension
Table 2 Accuracy and precision of ricin in serum matrix in LC-QqQMS/MS, with a quantification peptide T7A (n=3) Added (ng/mL)
Found (ng/mL)
Accuracya (%)
Precisionb (%)
10 50 80
7.8±0.38 36.5±5.2 70±3.4
78 73 87
4.8 14.2 5
a
Fig. 4 LC-QqQ-MS/MS chromatograms of blank serum (a) and ricin spiked at 2.5 ng/mL (b) and 5 ng/mL, (c) into blank serum, respectively
Expressed as [(mean found concentrations/added concentrations) × 100].
b
Expressed as the relative standard deviation (RSD).
X. Ma et al.
of incubation time to 16 h did not remarkably gain higher peptide yields than at 6 h. Method performance The method performance was evaluated by assessing the quantification results of ricin spiked in blank serum. For generating a calibration curve, purified ricin was serially diluted into blank serum to make samples with different concentrations, then these samples were incubated with mAb-MBs, and the captured ricin was eluted from MBs, followed by trypsin digestion and LC-QqQ-MS analysis. Peptide T7A was found to offer the best sensitivity and thus employed as the quantification peptide, with a limit of detection (LOD) of 2.5 ng/mL (S/N is greater than 3), and the limit of quantification (LOQ) was 5 ng/mL (S/N is greater than 10) (Fig. 4). Although this sensitivity was only comparable to the LOD of conventional ELISA, our method established herein could provide an unambiguous identification result at the same time. The calibration curve had a linear response from 5 to 80 ng/mL with a linear equation y=0.005x–0.014 (r2 =0.99, Fig. 5a). Serum samples containing 10, 20, and 50 ng/mL of ricin were analyzed in triplicate, and the accuracy values at low, middle, and high levels were 78 %, 73 %, and 87 %, respectively, with precision values ranging from 4.8 % to 14.2 % (Table 2). It was believed that the accuracy of the measurement could be further improved through isotope dilution mass spectrometry. Determination of residual ricin in serum samples of poisoned rats In case of terrorist attacks of ricin or accidental ingestion of castor beans, it is of great importance to analyze biomedical samples. To test the robustness of our method, the contents of residual ricin in rat serum were measured, which were collected at different times after i.v. administration of ricin to rats at a dosage level of 50 μg/kg (ca. 10 LD50). The trend of residual ricin over time was plotted as using peptide T7A for quantification (Fig. 5b). The results showed that both T7A and T14B (data not shown) appeared within 2 h after intoxication, and T7A could be detected up to 12 h. After that, the amount of residual ricin was lower than the LOD of this method. The determined amount of ricin in serum was much lower than that of administration because of the following reasons: a substantial part of ricin was delivered into various tissues via blood flow, and the antibody mAb 3D74 could only bind intact ricin, and some divided ricin A- or B-chains might be missed.
Conclusions The sensitive and quantitative LC-ESI-MS/MS methods were developed for identification and quantification of ricin in
serum samples. In this method, a magnetic immunocapture pretreatment based on specific mAb-coated MBs was implemented in order to extract and enrich target biotoxin from complicated matrices. Although comparing with current immuno-based bioassays, the LOD of this method was not quite superior, the accurate MS analysis of this work provided a great advantage of unambiguous identification because of the simultaneous monitoring of characteristic peptides of ricin. Therefore, ricin can be accurately distinguished from RCA120 through specific peptides T7A and T14B. Finally, the method was successfully employed to determine residual ricin in serum of poisoned rats. The residual ricin in serum of poisoned rats can be detected until 12 h later at a concentration of ca. 10 ng/mL. Acknowledgments This research was supported by grants from the National Natural Science Foundation of China (grant nos. 21175152, 81001262, and 81202248), and National Science and Technology Major Project of the Ministry of Science and Technology of China (grant no. 2012ZX09301003-001-010).
References 1. Barbieri L, Battelli MG, Stirpe F (1993) Ribosome-inactivating proteins from plants. Biochim Biophys Acta 1154:237–282 2. Lord JM, Roberts LM, Robertus JD (1994) Ricin: structure, mode of action, and some current applications. FASEB J 8:201–208 3. Papaloucas M, Papaloucas C, Stergioulas A (2008) Ricin and assassination of Georgi Markov Pakistan. J Biol Sci 11:2370–2371 4. Ma X, Liu H, Tang J, Guo L, Xie J (2011) Comparison of three enzyme-linked immunosorbent assay methods for quantitative determination of ricin. Chin J Anal Chem 39:685–689 5. Godal A, Olsnes S, Pihl A (1981) Radioimmunoassays of abrin and ricin in blood. J Toxicol Environ Health 8:409–417 6. Ligler FS, Taitt CR, Shriver-Lake LC, Sapsford KE, Shubin Y, Golden JP (2003) Array biosensor for detection of toxins. Anal Bioanal Chem 377:469–477 7. Goldman ER, Clapp AR, Anderson GP, Uyeda HT, Mauro JM, Medintz IL, Mattoussi H (2004) Multiplexed toxin analysis using four colors of quantum dot fluorore agents. Anal Chem 76:684–688 8. Araki F, Funatsu G (1987) The complete amino acid sequence of the B-chain of ricin E isolated from small-grain castor bean seeds. Biochim Biophys Acta 911:191–200 9. Duriez E, Fenaille F, Tabet JC, Lamourette P, Hilaire D, Becher F, Ezan E (2008) Detection of ricin in complex samples by immunocapture and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. J Proteome Res 7:4154–4163 10. McGrath SC, Schieltz DM, McWilliams LG, Pirkle JL, Barr JR (2011) Detection and quantification of ricin in beverages using isotope dilution tandem mass spectrometry. Anal Chem 83:2897–2905 11. Ostin A, Bergstrom T, Fredriksson SA, Nilsson C (2007) Solventassisted trypsin digestion of ricin for forensic identification by LCESI MS/MS. Anal Chem 79:6271–6278 12. Tang J, Xie J, Shao N, Yan Y (2006) The DNA aptamers that specifically recognize ricin toxin are selected by two in vitro selection methods. Electrophoresis 27:1303–1311 13. Skibeli V, Nissen-Lie G, Torjesen P (2001) Sugar profiling proves that human serum erythropoietin differs from recombinant human erythropoietin. Blood 98:3626–3634
Identification and quantification of ricin 14. Guan F, Uboh CE, Soma LR, Birks E, Chen J, Mitchell J, You Y, Rudy J, Xu F, Li X, Mbuy G (2007) LC-MS/MS method for confirmation of recombinant human erythropoietin and darbepoetin α in equine plasma. Anal Chem 79:4627–4635 15. Becher F, Duriez E, Volland H, Tabet JC, Ezan E (2007) Detection of functional ricin by immunoaffinity and liquid chromatography tandem mass spectrometry. Anal Chem 79:659–665 16. Guo L, Wei Y, Xie J (2011) Analysis of ricin: immunocapture. In: Vanninen P (ed) Recommended operating procedures for analysis in
the verification of chemical disarmament. University of Helsinki, Helsinki, pp 419–423 17. Men J, Lang L, Wang C, Wu J, Zhao Y, Jia P, Wei W, Wang Y (2010) Detection of residual toxin in tissues of ricin-poisoned mice by sandwich enzyme-linked immunosorbent assay and immunoprecipitation. Anal Biochem 401:211–216 18. Araki F, Funatsu G (1987) The complete amino acid sequence of the B-chain of ricin E isolated from small-grain castor bean seeds. Biochim Biophys Acta 911:562–568
