Article pubs.acs.org/ac
Liquid Chromatography, in Combination with a Quadrupole Time-ofFlight Instrument (LC QTOF), with Sequential Window Acquisition of All Theoretical Fragment-Ion Spectra (SWATH) Acquisition: Systematic Studies on Its Use for Screenings in Clinical and Forensic Toxicology and Comparison with Information-Dependent Acquisition (IDA) Andreas T. Roemmelt, Andrea E. Steuer, Michael Poetzsch, and Thomas Kraemer* Department of Forensic Pharmacology & Toxicology, Zurich Institute of Forensic Medicine, University of Zurich, Zurich, Switzerland S Supporting Information *
ABSTRACT: Forensic and clinical toxicological screening procedures are employing liquid chromatography−tandem mass spectrometry (LC-MS/MS) techniques with information-dependent acquisition (IDA) approaches more and more often. It is known that the complexity of a sample and the IDA settings might prevent important compounds from being triggered. Therefore, data-independent acquisition (DIA) methods should be more suitable for systematic toxicological analysis (STA). The DIA method sequential window acquisition of all theoretical fragment-ion spectra (SWATH), which uses Q1 windows of 20−35 Da for data-independent fragmentation, was systematically investigated for its suitability for STA. Quality of SWATHgenerated mass spectra were evaluated with regard to mass error, relative abundance of the fragments, and library hits. With the Q1 window set to 20−25 Da, several precursors pass Q1 at the same time and are fragmented, thus impairing the library search algorithms to a different extent: forward fit was less affected than reverse fit and purity fit. Mass error was not affected. The relative abundance of the fragments was concentration dependent for some analytes and was influenced by cofragmentation, especially of deuterated analogues. Also, the detection rate of IDA compared to SWATH was investigated in a forced coelution experiment (up to 20 analytes coeluting). Even using several different IDA settings, it was observed that IDA failed to trigger relevant compounds. Screening results of 382 authentic forensic cases revealed that SWATH’s detection rate was superior to IDA, which failed to trigger ∼10% of the analytes.
I
these LC-MS screenings are based on triple-quadrupole mass spectrometers employing a multiple reaction monitoring (MRM) survey scan, followed by a product ion scan.2−5 The resulting MS2 spectra can be used for unequivocal identification by comparison with library spectra. These methods are strictly targeted. For untargeted screenings, high-resolution mass spectrometry (HRMS) seems to be more suitable. In screenings, employing single-stage time-of-flight (TOF) MS identification of compounds is based on the accurate mass of
n forensic and clinical toxicology, the identification of as many licit or illicit drugs, poisons, and metabolites as possible in biological matrices (such as blood, urine, hair, nails, or tissue samples) is of major interest. Immunoassay methods are often used for prescreening purposes, but results must be confirmed by a second and independent methodusually via hyphenated techniques. Gas chromatography−mass spectrometry (GC-MS) screening has been the gold standard for systematic toxicological analysis (STA).1 In the past decade, liquid chromatography−mass spectrometry (LC-MS) methods have been gaining importance for that purpose, because of the advantages of easier sample preparation and accessibility of nonvolatile, polar, and thermally unstable compounds. Most of © 2014 American Chemical Society
Received: August 21, 2014 Accepted: October 20, 2014 Published: October 20, 2014 11742
dx.doi.org/10.1021/ac503144p | Anal. Chem. 2014, 86, 11742−11749
Analytical Chemistry
Article
dem, zolpidem d6 and zopiclone were purchased from LGC (Wesel, Germany). Working solutions of the analytes were prepared at concentrations of 100 μg/mL and 10 μg/mL in methanol or acetonitrile via dilution from the respective stock solution. HPLC and HRMS Settings. HPLC separation was performed using an UltiMate 3000 RSLC system (Thermo Fischer Scientific, San Jose, CA), configured in binary highpressure gradient mode and controlled by Chromeleon 6.80 software (Thermo Fischer Scientific). Mobile phase A consisted of 25 mM ammonium acetate buffer with 0.1% acetic acid, and mobile phase B consisted of acetonitrile with 0.1% acetic acid. The gradient was programmed as follows: 0.00−1.00 min, 5% eluent B; 1.01−5.00 min, gradient increase to 20% eluent B; 5.01−7.00 min, 22% eluent B; 7.01−9.00 min, 30% eluent B; 9.01−13.00 min, 40% eluent B; 13.01−14.00 min, 50% eluent B; 14.01−16.00 min, hold 50% eluent B; 16.01−17.00 min, 95% eluent B; 17.01−19.00 min, hold 95% eluent B; 19.01− 20.00 min, 5% eluent B. For re-equilibration of the HPLC column, the gradient was set to 5% eluent B for 5 min. The column oven was set at 40 °C, and the autosampler was cooled at 7 °C. The flow rate was 0.5 mL min−1. Ten microliters (10 μL) of the samples were injected onto a Synergi Polar RP column (100 mm × 2.0 mm inner diameter (I.D.); 2.5 μm particle size, 100A) (Phenomenex, Torrance, CA, USA) guarded with a C18 guard column (2.0 mm I.D. × 4.0 mm) (Phenomenex). Mass spectrometric detection was performed using a QTOF MS (TripleTOF 5600, ABSciex, Concord, Ontario, Canada) operated in positive ionization mode with a DuoSpray ion source at a resolving power (full width at half-maximum (fwhm) at m/z 400) of 30 000 in MS and 30 000 in MS2 (highresolution mode) or 15 000 (high-sensitivity mode). The automated calibration device system (CDS) was set to perform an external calibration every four samples. The source conditions were as follows: temperature, 500 °C; curtain gas (CUR), 25 psi; ion source gas (GS) 1 and 2 at 45 and 57 psi (laboratory frame), respectively; and ion-spray voltage floating (ISVF) at 5.5 kV. All MS parameters were controlled by AnalystTF Software 1.6 (ABSciex). Data were processed with PeakView 2.0 Software (ABSciex), MasterView 2.0 (ABSciex), and MultiQuant 2.1 software (ABSciex). The acquisition using SWATH consisted of a full scan, followed by a Q1 isolation strategy. The full scan covered a mass range of m/z 100−700 with a collision energy (CE) of 10 eV and an accumulation time of 100 ms. The Q1 isolation strategy covered a mass range of m/z 100−650 with a 20 Da window width for Q1 isolation (overlap 1 Da). Each SWATH window had an accumulation time of 40 ms, using a CE value of 35 eV with a collision energy spread of 15 eV in highresolution mode. The second acquisition method was IDA. The acquisition using IDA consisted of a full scan and information-dependent trigger events. The accumulation time for full scan was 100 ms for scanning a mass range from m/z 100 to m/z 700. The accumulation time for each IDA experiment was 50 ms, and the CE was set to 35 eV with a CE spread of 15 eV in highsensitivity mode. IDA criteria were as follows: 10 most intense ions (number of IDA experiments) with an intensity threshold above 100 cps, isotope exclusion was switched off, and an exclusion time of 6 s (half peak width) was set. Dynamic background subtraction was switched on.
the protonated/deprotonated molecular ion, isotopic pattern, and the a priori known retention time.6−13 The isotopic pattern of the analyte helped in reducing, but not in completely avoiding, false positive results.14 Analysis using a combination of a quadrupole and a TOF instrument (QTOF) allows for the selection of a precursor, which can subsequently be fragmented in a collision cell. Fragments are measured in the TOF section and generate a high-resolution product ion (PI) spectrum. The selection of the precursor is crucial and is based on the user’s predefined criteria (information-dependent acquisition, IDA).15−18 There is always a compromise between mutually contradictory settings. A large number of IDA experiments theoretically allows for more trigger events; however, it leads to longer cycle times. Long exclusion times help prevent multiple trigger events for one compound, but reduce the chance of positive identification when the actual peak maximum is excluded. Optimal settings for all situations seem to be impossible and, thus, the risk of missing relevant ions is always present when using IDA. For an actual comprehensive screening, data-independent (untargeted) acquisition (DIA) procedures are necessary. DIA methods offer the possibility of unrestricted and unbiased fragment ion spectra generation over the entire LC run with the disadvantage that the missing preselection may lead to impure mass spectra, which may be difficult to identify. In this study, a highly specific DIA method for QTOF instruments called sequential window acquisition of all theoretical fragment-ion spectra (SWATH) was systematically investigated for suitability for clinical and forensic toxicological screening. SWATH acquisition consists of a recurring cycle of a survey scan and a Q1 isolation strategy. In the first step, a survey scan with low collision energy covers the user-defined mass range (Q1 set to full transmission). The mass range then is consecutively scanned using predefined Q1 windows (typically 20 Da), applying a range of collision energies to produce product ion spectra. SWATH acquisition has already been used in proteomics research19,20 and in metabolomics.21 In the present study, the suitability of LC QTOF with SWATH acquisition for routine forensic screenings should be assessed. For that purpose, the quality of SWATH-generated mass spectra and the detection rate of IDA vs SWATH for small molecules was investigated. In addition, 382 authentic forensic cases were screened using SWATH, and results were compared with those generated by IDA on a QTOF.
■
EXPERIMENTAL SECTION Chemicals and Reagents. The chemicals and solvents for the high-performance liquid chromatography (HPLC) system were obtained from the following suppliers: water (LC-MS grade), acetonitrile (LC-MS grade), and acetic acid (99% purity, for analysis) were purchased from Merck (Darmstadt, Germany), and ammonium acetate was purchased from Sigma−Aldrich (Buchs, Switzerland). Agomelatine, AM 1241, JWH 018, JWH 020, and WIN 55,212-2 were purchased from Cayman Chemical (Ann Arbor, MI, USA). Amiodarone, amitriptyline, amitriptyline d3, amphetamine, amphetamine d6, citalopram, clomipramine, clomipramine d3, cocaine, cocaine d3, diazepam, diazepam d5, fentanyl, ketamine, levamisole, MDMA, methadone d9, methylphenidate, methylphenidate d9, morphine d3, morphine M (6 beta-D-glucuronide), quetiapine, reboxetine, trazodone, trimipramine, vardenafil, venlafaxine-M (O-desmethyl-), zolpi11743
dx.doi.org/10.1021/ac503144p | Anal. Chem. 2014, 86, 11742−11749
Analytical Chemistry
Article
Blood Samples and Sample Preparation. Blank blood samples for the IDA vs SWATH comparison were obtained from volunteers of the Zurich Institute of Forensic Medicine (ZIFM) of Zurich University, Switzerland. All samples were routinely prescreened. After anonymization, 382 authentic blood samples from routine analysis of the ZIFM were used for comparison experiments: 48 originated from post-mortem cases, and the remainding samples were from other cases, such as driving under the influence of drugs (DUID). Two hundred microliters (200 μL) of whole blood were precipitated with 600 μL of ice cold acetonitrile and shaken for 10 min. After centrifugation for 10 min at 12 000 rpm, the supernatant was evaporated to dryness under a stream of nitrogen at 50 °C and reconstituted in 50 μL of a mixture of eluent A and B (50:50; V/V). Aliquots of 10 μL were injected into the liquid chromatography−high-resolution mass spectrometry (LC-HRMS) system. Identification of the Compounds and Library Settings. Identification of the compounds was always the same, except for the coelution experiment for IDA vs SWATH (cf. Assessment Criteria for Coelution Experiment for IDA vs SWATH). For the identification of the compounds, MasterView 2.0, which is integrated in the PeakView 2.0 software, was used. A total of 1326 compounds and their empirical formulas were integrated in an intact accurate mass list (XIC list). A mass error of the precursor of 60 (library score points), still tentative above 25, and negative at