Validation of the Mass-Extraction-Window for ... - ACS Publications

Feb 2, 2016 - Quantitative Mass Spectrometry Facility, University Hospital of Lausanne, CHUV, 1011 Lausanne, Switzerland. §. Institut de Police Scien...
0 downloads 0 Views 805KB Size
Subscriber access provided by Weizmann Institute of Science

Article

Validation of the mass-extraction-window for quantitative methods using liquid chromatography high resolution mass spectrometry. Gaétan Glauser, Baptiste Grund, Anne-Laure Gassner, Laure Menin, Hugues Henry, Maciej Bromirski, Frederic Schutz, Justin McMullen, and Bertrand Rochat Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.5b04689 • Publication Date (Web): 02 Feb 2016 Downloaded from http://pubs.acs.org on February 22, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Validation of the mass-extraction-window for quantitative methods using liquid chromatography high resolution mass spectrometry.

Authors : Gaétan Glauser (1), Baptiste Grund (2), Anne-Laure Gassner (3), Laure Menin (4), Hugues Henry (5), Maciej Bromirski (6), Frédéric Schutz(7), Justin McMullen (2), Bertrand Rochat (2)*

Affiliation: (1)

Neuchâtel Platform of Analytical Chemistry, Institut de Chimie, Université de Neuchâtel, Switzerland,

(2)

Quantitative Mass Spectrometry Facility, University Hospital of Lausanne; CHUV, 1011 Lausanne, Switzerland,

(3)

Institut de Police Scientifique, University of Lausanne, Batochime, 1015 Lausanne, Switzerland

(4)

EPFL, Institut of chemical sciences and engineering, Batochime, 1015 Lausanne, Switzerland

(5)

BioID, Department of Laboratories, University Hospital of Lausanne, CHUV; Lausanne, Switzerland

(6)

Thermo Fisher Scientific, Bremen, Germany;

(7)

Swiss Institute of Bioinformatics; Génopode, University of Lausanne 1015 Lausanne, Switzerland

*Corresponding author : Bertrand Rochat, PhD, Quantitative Mass Spectrometry Facility [qMSF] BH18-228; CHUV [Centre Hospitalier Universitaire Vaudois] Rte du Bugnon 46, CH - 1011 Lausanne, Switzerland e-mail: [email protected] ; phone: + 41 21 314 41 58; fax : + 41 21 314 42 88.

Short title: Validation criteria for quantitative HRMS analyses. Key words: mass-extraction-window, mass accuracy, high-resolution mass-spectrometry, quantification, validation.

1 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 26

ABSTRACT A paradigm shift is underway in the field of quantitative liquid chromatography-mass spectrometry (LCMS) analysis thanks to the arrival of recent high-resolution mass spectrometers (HRMS). The capability of HRMS to perform sensitive and reliable quantifications of a large variety of analytes in HR-full scan mode is showing that it is now realistic to perform quantitative and qualitative analysis with the same instrument. Moreover, HR-full scan acquisition offers a global view of sample extracts and allows retrospective investigations as virtually all ionized compounds are detected with a high sensitivity. In time, the versatility of HRMS together with the increasing need for relative quantification of hundreds of endogenous metabolites should promote a shift from triple-quadrupole MS to HRMS. However a current “pitfall” in quantitative LC-HRMS analysis is the lack of HRMS-specific guidance for validated quantitative analyses. Indeed, false positive and false negative HRMS detections are rare, albeit possible, if inadequate parameters are used. Here we investigated two key parameters for the validation of LC-HRMS quantitative analyses: the mass accuracy (MA) and the mass-extraction-window (MEW) that is used to construct the extracted-ion-chromatograms. We propose MA-parameters, graphs and equations to calculate rational MEW width for the validation of quantitative LC-HRMS methods. MA measurements were performed on four different LC-HRMS platforms. Experimentally determined MEW values ranged between 5.6 to 16.5 ppm and depended on the HRMS platform, its working environment, the calibration procedure and the analyte considered. The proposed procedure provides a fit-for-purpose MEW determination and prevents false detections.

2 ACS Paragon Plus Environment

Page 3 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

INTRODUCTION Over the last decades, various guidelines have been edited by Authorities for the validation of quantitative LC-MS methods and analytical sequences with intra- and inter-day assays.101,102,1 Thus, quantitative LCMS analyses are performed following strict rules in laboratories producing certified data. These guidelines can be applied to triple-quadrupole (QQQ-MS) ion trap, quadrupole-time-of-flight (QTOF) and Orbitrap (Orbi-; Orbitrap® and (Q)Exactive®) mass analyzers. In general, these guidelines refer to: 1. Selectivity, 2. Accuracy, precision and recovery, 3. Calibration curve, 4. Sensitivity, 5. Reproducibility and 6. Stability.101,102 Various other good practices are left to the bioanalyst’s discretion, for instance: MS system suitability, instrument calibration and maintenance, possible interference evaluation, loss by adsorption on surface etc. Standard operating procedures (SOPs) have been written in most labs that perform quantitative analysis using QQQ-MS. Moreover, the way to perform LC-MS analysis using QQQ-MS and to treat the data, are well established for analytical chemists. Thus, logically, it may be challenging to change “old habits”, SOPs and workflows and to switch from QQQ-MS to another MS technology. However, accumulating data show that recent high-resolution instruments (HRMS) provide similar quantitative performance to QQQ-MS.2,3,4,5,6,7,8,9 LC-HRMS analysis usually employs HR-full scan (HRFS), which records all ionized compounds in a global acquisition, without the need for setting selective reaction monitoring (SRM) and optimizing collision energies. This is a clear advantage of HRMS over QQQ-MS.10,11,12,13,14,15,16,17 In HR-FS acquisition, the compound is depicted by the construction of an extracted-ion-chromatogram (XIC) using a mass-extraction-window (MEW) centered on the theoretical m/z (m/ztheor). Detection depends on the analyte m/z in the biomatrix, the mass resolution (R), the mass accuracy (MA; also called mass deviation or mass shift) and the MEW. Calculations of R and MA, as well as a depiction of MEW, are 3 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 26

shown in Figure S-1, Supporting Information. R, MA and MEW are critical parameters for m/z extraction. Indeed, there is a risk of false positive detections because of unresolved isobaric interferences and/or too broad MEW set for the XIC construction (Figure S-2, Supporting Information). Alternatively, there is a risk of false negative detections that are the result of a MA shift that moves the m/z measured (m/zmeas) outside the MEW (Figure S-2, Supporting Information). False positive and false negative detections can also occur with QQQ-MS.18,19,20 However, they are rare and, for this reason, are not considered as a problem in SRM mode. From our experience, as well as data reported elsewhere,20,21 false positive and false negative detections are also rare in LC-HRMS analyses using HR-FS mode so long as appropriate R and MEW width are used and acceptable mass shifts are observed. Thus, there is a need to validate the MEW value in quantitative LC-HRMS methods. In this study, based on MA measurements regrouped in an overall assay, we propose novel MAparameters, graphs and equations to calculate rational MEW for the validation of quantitative LC-HRMS methods. Concrete examples using data recorded on 2 Orbi- (Q Exactive®) and 2 Q-TOF MS systems are provided.

EXPERIMENTAL SECTION LC-HRMS systems and parameters. Various chromatographic systems have been used but all consisted of a (U)HPLC pump and an autosampler maintaining injection vials at 10-15°C. Various samples were extracted or prepared and injected onto various C18 analytical columns (30 to 50) x 2.1 mm (L x i.d.). The mobile phase was composed of (a) 0.1 to 1% formic acid (FA) and (b) acetonitrile (MeCN) or methanol (MeOH) with 0.1 to 1% FA, delivered at 0.3 to 0.5ml/min. The columns were coupled to electrospray

4 ACS Paragon Plus Environment

Page 5 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

sources (ESI) and HRMS systems. Total run time was 10 to 20 min. For more details, refer to the Results section. The LC systems were coupled to a Q Exactive Focus, a Q Exactive-MS (Orbi-MS#1 and #2, respectively; Thermo, Germany), an Agilent 6530 Q-TOF-MS (Q-TOF#1, Agilent, USA) or a Synapt G2 Q-TOF-MS (Q-TOF#2, Waters, USA). Standard ESI parameters, in positive or negative mode (ESI+ and ESI-, respectively) were used. MS full scans were acquired over various mass ranges. For HRMS measurements, mass calibration was external (Orbi-MS#2 or Q-TOF#1; prior the analytical sequence) or internal with a lock-mass recalibrating the MS from 0 to 0.25 min in each run (Orbi-MS#1) or constantly over the entire analytical run (LocksprayTM on Q-TOF#2). Mass resolutions (R) were set at 70,000 at m/z = 200, 140,000 at m/z = 200, 15,000 at m/z = 300 and 20,000 at m/z = 500 for Orbi-MS#1, Orbi-MS#2, Q-TOF#1 and QTOF#2, respectively. LC-HRMS data acquisition, peak integration, and quantification were performed using Xcalibur, Masshunter or MassLynx softwares (Thermo, Agilent and Waters, USA, respectively). Extracted-ionchromatograms (XIC) were based on mass-extraction-windows (MEW) centered on m/ztheor of the drug considered. Mass accuracy was determined for each sample either with a contaminant ion (mean of ≥ 15 scans) found in the mobile phase (polysiloxanes (C2H6SiO)6 at m/z = 445.12003 or (C2H6SiO)5 at m/z = 371.10123, plasticizers C24H38O4 at m/z = 391.28429 or C10H15NO2S at m/z 214.08963; N,N'Dicyclohexylurea C13H24N2O at m/z = 225.19614 in ESI+ or C3H5O3 at m/z = 182.9885 in ESI-) or with the quantified analytes (mean of ≥ 15 scans at 10-20% of the analyte peak height). Sample biomatrices and preparations. For Orbi-MS #1 and #2 instruments, human plasma samples were prepared from whole blood withdrawn in collection tubes (monovettes with EDTA anticoagulant). For Orbi-MS#1, tyrosine kinase inhibitors (e.g. nilotinib) were extracted by protein precipitation using MeCN

5 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 26

(plasma: MeCN, 1:3, v:v), centrifuged at 20,000g for 15 min at 4°C and supernatants were removed and diluted with aqueous solution and placed in injection vials.22 For Orbi-MS #2, plasma samples were extracted by solid-phase-extraction. Eluates were dried and reconstituted with 80µL of methanol: ammonium formate 20mM 1:2, v:v.23 For Q-TOF#1 instrument, samples were prepared from stock solutions of organic gunshot residue standards and diluted to desired concentration in water. For Q-TOF#2 instrument, samples were prepared from Arabidopsis or maize plants according to established protocols.8,24,25

GRAPHS, MA-PARAMETERS AND MEW DETERMINATION WITH EQUATIONS. In order to avoid false positive and false negative detections in LC-HRMS (see Figure S-2, Supporting Information), it is essential to validate MEW width. For this purpose, we propose to make 2 graphs and to determine MA-parameters in order to determine rational MEW width based on the equations proposed below. Evaluation of false negative and positive detections by drawing 2 graphs. The graphs are drawn from the analysis of ≥ 3 matrix extracts from ≥ 3 different biological replicates containing the analyte(s) at 1 to 5x the lowest limit of quantification (LLOQ). The first plot (Graph #1) depicts LC-HRMS peak area of the analyte(s) at various MEW widths (Figures 1A). The second plot (Graph #2) shows the number of LCHRMS peaks around the analyte retention time against various MEW widths (Figure 1B). A third plot (optional) representing XICs centered on the analyte m/z at various MEW widths could be built based on a post-column infusion of the analytes(s) using a Tee connector. The analyte signal intensity should be approximately 10 to 100x the noise generated by a blank chromatogram. This chromatographic plot has been proposed by Kaufmann et al., to evaluate mass shifts as the consequence of unresolved isobaric

6 ACS Paragon Plus Environment

Page 7 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

interferences or ion coalescence from the injected extracts (see examples in Figure S-3A/B, Supporting Information).26,27 Typically, the graphs should be drawn from XIC built with MEW of ± 1, 2.5, 5, 10, 25, 50, 100 and 200 ppm (or corresponding values in mDa). Critical MEW width(s) for which false detections are observed will be determined. Eventually, the MEW value, for which a risk of false positive or negative detections exists, will be established for a given analyte m/z in its biomatrix extract. Determination of appropriate MEW width with MA-parameters and equations. To determine rational MEW width, the MA-parameters investigated in LC-HRMS analyses are R, MA mean (MAmean), MA precision (MAprecision) and MA interval (MAinterval), and are depicted in Figure 2A. These MA-parameters are determined in MA-related intra-assays (MAintra-assay) that take into account the following influences: I) long period of time (≥ month), II) individual scan variability through the analyte peak, III) calibrant levels and IV) biomatrix origin (Figure 2B). The greatest (most deviant) MAmean, MAprecision and MAinterval values found in all MA-related intra-assays are chosen for the overall assay and give the MAoverall parameters (Figure 2B). From these MA parameters, rational and appropriate MEW widths are calculated according to specific equations given thereafter. Resolution (R): •

Equation #1: R = m/mFWHM

where m is the analyte m/z and mFWHM is the delta of m/z distribution (in Da or u) at Full-Width-HalfMaximum (corresponding to 50% of the maximum peak height; see Figure S-1, Supporting Information 1). R is dependent on m/z and should be given at the analyte m/z (Ranalyte). MA-parameters:

7 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



Page 8 of 26

Equation #2: MA = [(m/zmeas - m/ztheor) / m/ztheor] x 106, in [ppm] or m/zmeas - m/ztheor, in [Da], where m/zmeas is the mean of individual m/z measured across the LC-MS peak and m/ztheor is the theoretical m/z corresponding to a specific chemical composition (e.g. monoisotopic mass + H+).



MAmean is the average of MA values in an intra-assay (MAmean

intra-assay)

or across all intra-assays

(MAmean overall). See Figure 2. •

MAprecision is the standard deviation of MA values in an intra-assay (MAprecision intra-assay) or across all intra-assays (MAprecision overall). See Figure 2.



MAinterval is the greatest MA interval in an intra-assay (MAinterval

intra-assay)

or across all intra-assays

(MAinterval overall). See Figure 2. MAinterval takes into account the MAlowest and MAhighest as the lowest and highest MA individual values and is calculated with the following equation: •

Equation #3: MAinterval = │(MAlowest - MAhighest)│

It should be noted that MA-parameters were previously called differently28 but terminologies have been changed for the sake of clarity in the present work. MA-related intra-assays. We propose to perform 2 types of MA-related intra-assays to determine the MA-parameters (MAmean, MAprecision and MAinterval). The first type of intra-assays (I) is independent of the method to be validated and takes into account MA over a long period of time (MAover-time). It represents the MA of a specific HRMS instrument in its working environment (mass calibration procedure, selected resolution, etc.) and should be considered as a system suitability test for MA shifts. MAover-time intra-assays can be used for further method validations and are performed as follows: I. MAover-time. MA values are determined in each sample of a long analytical sequence (typically 6 to 24h) using any m/z ion at any retention time, which is representative of MA. A mean of ≥ 15 scans should be used. The ion can be a real peak or a contaminant found in the mobile phase. At least 5 MAover-time assays

8 ACS Paragon Plus Environment

Page 9 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

should be recorded over a long period of time (≥ 1 month). MAmean, MAprecision and MAinterval are calculated in these assays (Figure 2A). The second type of assays (IIa, b and c) is related to the analyte to be quantified and seeks to decipher the different factors which may influence mass accuracy: individual scans through the analyte peak (MAinterscan),

calibrant levels (MAcalibration) and biomatrix origin (MAbiomatrix) (Figure 2B). MAmean, MAprecision and

MAinterval are calculated for each of these factors and can be obtained from one analytical sequence (Figure 2A). MAinter-scan, MAcalibration and MAbiomatrix parameters in these intra-assays are performed as follows: IIa. MAcalibration. MA values are determined in all calibration samples (MAcalibration), from the lowest to the upper limit of quantification (LLOQ and ULOQ respectively) and across the analyte peak measured at 1020% of the peak height (mean of ≥ 15 scans). MAcalibration should be determined with the analyte m/z in the matrix extract. IIb. MAbiomatrix. MA are measured across the analyte peak (MAbiomatrix), at 10-20% of the peak height (mean of ≥ 15 scans) and in extracts from ≥ 5 different biological samples. In order to reveal possible interferences, MAbiomatrix should be determined at 1 to 5x the LLOQ levels. IIc. MAinter-scan. MA values are determined in each individual scan across the analyte peak (MAinter-scan) and are measured at 1 to 5x the LLOQ calibrants in the matrix extract and at least in duplicate. Depending on the type of HRMS instrument, individual scan variability can significantly exceed the average MA. The greatest MAmean, MAprecision and MAinterval values found in all MA-related intra-assays (I, IIa-c) are chosen and regrouped in the overall assay resulting in 3 MAoverall parameters (see Figure 2B).

9 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 26

Determination of MEW values. Using the following equations, a rational calculation of the appropriate mass-extraction-window (MEW) can be made. According to m/z (m) of the analyte(s), Ranalyte, established MAmean, MAprecision and MAinterval values, the MEW widths are determined as follows: • Equation #4: MEWintra-assay (in Da or ppm) = ± (│MAmean intra-assay │+ 4x MAprecision intra-assay) MEWintra-assay is the MEW value determined with all samples of an intra-assay and used for a single intra-assay. • Equation #5: MEWoverall (in Da or ppm) = ± (│MAmean overall │+ MAinterval overall) MEWoverall is the MEW value given for the method and used for all further analytical sequences. • Equation #6: MEWresolution (in ppm) = ± (1/Ranalyte x 106 + │MAmean overall │+ MAinterval overall) • Equation #7: MEWresolution (in Da) = ± (m/Ranalyte + │MAmean overall│+ MAinterval overall) Where “m” is the analyte theoretical m/z. MEWresolution was previously denominated differently28 but was renamed for clarification. MEWresolution corresponds to the mass window in which the analyte that would show a shift of MA as the results of 2 merging ions, would still be detected, whereas a narrower MEW could result in a false negative detection. In the case where MAmean overall and MAinterval overall would be null in equation #6, MEWresolution would be ± 100, 50, 25, 14, 7 and 4 ppm for resolutions, R, of 10k, 20k, 40k, 70k, 140k and 240k (realistic resolutions in actual HRMS instruments). Equations #4 and #5 consider the extraction of the centroid peak associated to the Gaussian m/z distribution (Figure S-1, Supporting Information). Equations #6 and #7 consider the complete profile distribution of m/z and are therefore related to the mass resolution at the analyte m/z (Ranalyte).

10 ACS Paragon Plus Environment

Page 11 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

MAprecision overall is not used in the determination of any proposed MEW. However, 4x MAprecision overall should be similar to MAinterval overall values (≤ 2 fold difference). If [(MAinterval overall) / (4x MAprecision overall)] ratio is > 2, false detections should be suspected as the results of coeluting isobaric interferences. In this case, the identity of the analyte showing MAhighest or MAlowest values (outliers) should be verified. This ratio should be determined in the validation of a LC-HRMS method. A chart which summarizes the successive steps in the determination of the possible false detections, the MA-parameters and MEW values, is provided in Figure 3.

RESULTS AND DISCUSSION Determination of MA-parameters and MEW values with four HRMS instruments. To illustrate the applicability of MEW calculation, MA-parameters were determined in MA-related intra-assays performed on 4 different LC-HRMS systems. It has to be noted that different analytical conditions and calibration procedures were applied on each HRMS platform and that MA was calculated on different samples and m/z. Therefore no strict comparisons should be made between the 4 analytical platforms. Figure 4 and Table 1 show the complete data and parameter determinations for all MA-related intra-assays performed with the analysis of etravirin drug on the Orbi-MS#1. The effects of the analyte concentration (calibration), individual scans at the LLOQ levels in plasma extracts (2.5 ng/mL) and biomatrix origin on MA were investigated. Individual scans, concentrations and biomatrices had no significant impact on MA (Figure 4). Figure S-4 of the Supporting Information shows MA-parameters determined in over-time and calibration intra-assays performed on the 3 other HRMS systems.

11 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 26

Figure S-5 of the Supporting Information shows MA of each individual scan across the analyte peak for 3 out of the 4 HRMS systems. Indeed, MEW value must also take into account MA in each individual scan across the LC-MS peak and not only the MA mean. In general, MAinterval

inter-scan

is greater at lower

intensities (Figure S-5, Supporting Information). Therefore, MA-parameters should be determined at 1 to 5x the LLOQ in the inter-scan intra-assay. By selecting a MEW value that is ≥ MAinterval in the inter-scan intra-assay, no reduced peak area resulting from a false negative scan should occur. Table 2 gives the MA-parameters and MEW related parameters for the 4 HRMS systems. Figure S-6 of the Supporting Information displays all detailed values. In Table 2, MAinterval

overall

is very close to 4x

MAprecision overall values for all HRMS systems tested (MAinterval overall = 0.8 to 1.2 of the [4x MAprecision overall] value). This indicates that no false detections (e.g. one outlier MA value) should be suspected. Selecting appropriate MEW width. Numerous quantitative LC-HRMS methods using Orbi- and Q-TOFMS have been already validated for various kinds of molecules in various biomatrices and with different sample preparations.29,30,31,32,33,34 These articles and their authors underscore the capabilities of HRMS to perform quantifications with comparable robustness, selectivity and sensitivity to QQQ-MS. For most validations, only the actual guidelines edited by Authorities are considered.101,102,1 A few attempts to define HRMS-specific criteria have been made and focused on the risk of false negative detections in screening analysis.28,35,36 Various recommendations have been proposed for the determination of MEW in relation to R. For instance, Vergeynst et al.36 proposed to use a MEW of ± 100, 50, 20 and 10 ppm for a resolution of 10,000, 20,000, 50,000 and 100,000, respectively. However, even if a mass shift can be the result of 2 merging unresolved ions because of insufficient resolution, MA deviation is not directly and not only correlated to the resolution, but rather to the calibration and MA accuracy of the HRMS during the analyte detection. Therefore, in order to limit the risk of false detection in quantitative analysis, MEW should take into account MAmean, MAprecision, MAinterval in the proposed MA-related intra-assays (inter-scan, biomatrix,

12 ACS Paragon Plus Environment

Page 13 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

calibration and over-time) and be calculated according to equations #4-7 presented above (See Figure 2). According to our equations, the following trends are observed for MEW values: MEWintra-assay ≤ MEWoverall ≤ MEWresolution. The various calculations give certain flexibility to bioanalysts to choose MEW widths with the appropriate tolerance. On the basis of our calculations of MEWintra-assay, MEWoverall and MEWresolution values for the 4 different HRMS instruments (Table 1, 2 and Figure S-6, Supporting Information), we show that in routine analyses, the MEWintra-assay and MEWoverall values are compatible with selective determinations (MEWoverall between ± 6 and 17 ppm) and reliable quantifications. Applying MEWintra-assay or MEWoverall in several quantitative applications did not lead to any false positive or negative detections. However, additional peaks or increased peak area may appear with MEW ≥ ± 20 ppm as it can be seen with nilotinib drug (Graph #2, Figure 1B). This may be the case if MEWresolution is chosen to build XIC chromatograms (Table 2). MEWresolution gives the broadest MEW value (Table 1 and 2) and is close to Vergeynst’s proposal, based solely on R (e.g. MEW = ± 50 and 10 ppm for R = 20,000 and 100,000, respectively).36 However, the above equation takes into account (│MAmean overall│ + MAinterval overall) in addition to Ranalyte. When Ranalyte becomes greater than 70,000, (│MAmean overall│ + MAinterval overall) increases MEWresolution substantially and becomes more and more important in inverse proportion to R (Figure S-7, Supporting Information). This justifies to add (│MAmean overall│ and MAinterval overall) in the calculation of MEWresolution. The question to know which R and MEW values are needed for selective detections in LC-HRMS analyses, has been investigated in a few studies.5,20,21,37,38 Data reveal that the selectivity depends on the biomatrix and its clean-up, the m/z value considered, the chromatographic conditions, the concentration of the analyte (S/N), R and MEW. These articles show that, when R is between 20,000 and 50,000 and XIC are constructed with a MEW of 5 to 10 ppm, HR-FS acquisition is as selective as low resolution SRM. When R and MEW are ≥ 50,000 and 5 to 10 ppm, respectively, HR-FS is even more selective than SRM

13 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 26

performed on QQQ-MS. On the contrary, a R of 10,000 is not selective enough in many cases, especially when low detection limit is expected.5,26,37 Noteworthy, various reports18,19,20,37 show that false positive interferences with low resolution SRM acquisitions can be solved with HRMS performing HR-FS acquisitions with Ranalyte ≥ 20,000 and a MEW ≤ ± 50 ppm. Based on these considerations and in our view, MEWoverall should be preferred as the best compromise between selectivity and risk of false detections. In addition, in our 4 MA-assays performed with 4 different HRMS instruments, (MAinterval overall) / (4x MAprecision overall) ratios were < 2 underscoring that there were no significant shifts of MA in any sample that could result in false detections.

CONCLUSION We have proposed an original procedure (Figure 3) to establish rational MEW value(s) in the validation of LC-HRMS quantitative methods. According to the determination of MEWoverall (found to be between ± 5.6 and 16.5 ppm in 4 different LC-HRMS set-up and methods), false negative detections in quantitative analysis using LC-HRMS are prevented whereas detection can be considered as selective as SRM performed on QQQ-MS. This level of selectivity can be obtained in HR-FS acquisition, which in parallel allows an in-depth, untargeted analysis of all metabolites in the sample. Indeed, the capacity to perform quantitative analyses with HR-FS acquisition opens the door to simultaneous quantitative and qualitative metabolite profiling and is a clear advantage in retrospective data mining for research or legal investigations.

14 ACS Paragon Plus Environment

Page 15 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Acknowledgement. Anne-Laure Gassner wishes to acknowledge the Swiss National Foundation (noPP00P1_150742) for financial support. We thank Drs. Anne Boddecs, Anne Cissencco and Denis Derume for their help.

Legend of Figures and Tables Figure 1. Graphs evaluating the risks of false positive detections in relation to the MEW width. A) Graph #1 built from a LC-HRMS analysis of a plant maize extract containing the analyte, chlorogenic acid at m/zmeas 353.0876. The analysis was performed on the Q-TOF#2 at R = 20’000 in centroid mode. Peak area of chlorogenic acid is plotted against MEW values and expressed relatively to the area obtained with MEW = ± 28 ppm (corresponding to ±10 mDa). MEW values < ± 7 ppm show reduced peak areas, while MEW values > ± 104 ppm increase peak areas as the result of an interfering peak, coumaroyl-hydroxycitric acid at m/zmeas 353.0508, at identical retention time. B) Graph #2 built from a LC-HRMS analysis of a plasma extract spiked with nilotinib drug (m/zmeas at 530.19197). Number of additional peaks is plotted against MEW values with a ± 2 min window around the drug retention time. The analysis was performed on the Orbi-MS#1. Figure 2. Determinations of MA parameters: MAmean, MAlowest, MAhighest, MAinterval and MAprecision in a LCHRMS analysis. MA-parameters can be determined in 2 types of MA-related intra-assays: (I) in samples over a long period of time, typically 6 to 24 h (MAover-time; x axis = time [h]) IIa. in calibration samples (MAcalibration; x axis = concentrations [ng/mL]), IIb. in biomatrices from different biological samples (MAbiomatrix; x axis = biomatrix number [N]) and IIc. individual scans of the LC-MS peak (MAinter-scan; x axis = scan number [N]). On the right-hand side, the intra-assays and the overall assay regrouping all MAparameters determined in the intra-assays are depicted. Type I intra-assay relates to the HRMS instrument

15 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 26

in its working environment and is a system suitability for MA shift. It can be used for all methods. Type II intra-assays are method and analyte-dependent. Figure 3. A flow chart representation of the complete procedure for the determination of false detections, MA-parameters in the MA-related intra-assays and the overall assay for the calculation of MEW widths. The black dot refers to the mass deviation related to the HRMS instrument in its working environment (I). It serves as system suitability for MA shift and can be used in all method validations. Stars depict the assays/graphs that are dependent on the LC-HRMS analysis and the analyte m/z and that have to be performed for each method validation (II). Figure 4. MA-related intra-assays for MA-parameter determination. MA deviation depicted in LC-HRMS assays performed on the Orbi-MS#1 (MAintra-assay): MAover-time, MAbiomatrix, MAcalibration and MAinter-scan. MAover-time was determined using 50 scans at a retention time of 2 minutes on a known plasticizer contaminant ion (C24H38O4 at m/ztheor = 391.28429). MAbiomatrix and MAcalibration were calculated using the analyte, etravirin drug, at m/ztheor 435.05635 (mean of ≥ 15 scans as one data point). MAinter-scan was calculated with 24 individual scans. For each sample, mass calibration was done from 0 to 0.25 min using the contaminant ion at m/ztheor = 214.08963. [N] stands for the number of data points. Table 1. Complete dataset obtained from MA-related intra-assays on the Orbi-MS#1 for the determination of MA-parameters and MEW values. MA-related intra-assays #1 to #5 are MAover-time assays (*, OT). The greatest values within the 8 intra-assays are selected for the determination of MA-parameters of the overall assay. Using these parameters and the equations given in the main text, MEW values are calculated (R at analyte m/z = 50,000). Table 2. MA-parameters in the overall assay and MEW values determined on the 4 HRMS instruments. Resolution at the analyte m/z and the type of mass calibration for each HRMS system are given. The 16 ACS Paragon Plus Environment

Page 17 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

complete dataset obtained for MA parameters on the 4 HRMS are given in the Figure S-1 (Supporting Information).

Supporting Information. Supporting Information S-1. Representation of m/z distribution and definition of R and MA. Supporting Information S-2. XICs of various analytes at 3 different MEWs from an LC-HRMS analysis. Supporting Information S-3. False negative and false positive detections. A) Post-column infusion. B) Example of a MA shift as the result of ion coalescence. Supporting Information S-4. MA deviation depending on time and concentration on 3 different LCHRMS platforms. Supporting Information S-5. Mass accuracy of each individual scan across LC-HRMS peaks. Supporting Information S-6. All individual MA parameter values of the MA-related intra- and interassays. Supporting Information S-7. MEWresolution values against R.

17 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 26

References 1.

Shah, V.P.; Bansal, S. Bioanalysis 2011, 3, 823-827.

2.

Kaufmann, A.; Butcher, P.; Maden, K.; Walker, S.; Widmer, M. Rapid Commun. Mass Spectrom. 2011, 25, 979-992.

3.

Henry, H.; Sobhi, H.R.; Scheibner, O.; Bromirski, M.; Nimkar, S.B.; Rochat, B. Rapid Commun. Mass Spectrom. 2012, 26, 499-509.

4.

Zhang, N.R.; Yu, S.; Tiller, P.; Yeh, S.; Mahan, E.; Emary, W.B. Rapid Commun. Mass Spectrom. 2009, 23, 1085-1094.

5.

Kellmann, M.; Muenster, H.; Zomer, P.; Mol, J.G.J. J. Am. Soc. Mass Spectrom. 2009, 20, 1464-1476.

6.

Morin, L.P.; Mess, J.N.; Garofolo, F. Bioanalysis 2013, 5, 1181-1193.

7.

Fung, E.N.; Xia, Y.Q.; Aubry, A.F.; Zeng, J.; Olah, T.; Jemal, M. J. Chromatogr. B 2013, 879, 2919-2927.

8.

Glauser, G.; Schweizer, F.; Turlings, T.C.; Reymond, P. Phytochem. Anal. 2012, 23, 520-528.

9.

Fedorova, G.; Randak, T.; Lindberg, R.H.; Grabic, R. Rapid Commun. Mass Spectrom. 2013, 27, 1751-1762.

10. Rochat, B.; Dahmane, E.; Zaman, K.; Csajka, C. Ther. Drug Monit. 2015, 37, 141-146. 11. Ramanathan, R.; Jemal, M.; Ramagiri, S.; Xia, Y.Q.; Humpreys, W.G.; Olah, T.; Korfmacher, W.A. J. Mass Spectrom. 2011, 46, 595-601. 12. Rochat, B. Bioanalysis 2012, 4, 1709-1711. 13. Josephs, J.L. Bioanalysis 2012, 4, 471-476. 14. Kaufmann, A.; Dvorak, V.; Crüzer, C.; Butcher, P.; Maden, K.; Walker, S.; Widmer, M.; Schürmann, A. J. AOAC Int. 2012, 95, 528-548. 15. Huang, M.Q.; Lin, Z.J.; Weng, N. Bioanalysis 2013, 5; 1269-76. 16. Rochat, B. Bioanalysis 2015, 7, 5-8. 17. Rochat, B.; Favre, A.; Sottas, P.E. Bioanalysis 2013, 5, 1149-1152. 18. Furlong, M.; Bessire, A.; Song, W.; Huntington, C.; Groeber, E. Rapid Commun. Mass Spectrom. 2010, 24, 1902-1910. 19. Kumar, P.; Rúbies, A.; Centrich, F.; Companyó, R. Meat Sci. 2014, 97, 214-219. 20. Kaufmann, A.; Butcher, P.; Maden, K.; Walker, S.; Widmer, M. Anal. Chim. Acta. 2010, 673, 60-72. 21. Kaufmann, A.; Butcher, P.; Maden, K.; Walker, S.; Widmer, M. Anal. Chim. Acta. 2015, 856, 54-67. 22. Haouala, A.; Zanolari, B.; Rochat, B.; Montemurro, M.; Zaman, K.; Duchosal, M.A.; Ris, H.B.; Leyvraz, S.; Widmer, N.; Decosterd, L.A. J. Chromatogr. B 2009, 877, 1982-1996. 23. Bermon, S.; Garnier, P.Y.; Hirschberg, A.L.; Robinson, N.; Giraud, S.; Nicoli, R.; Baume, N.; Saugy, M.; Fénichel, P.; Bruce, S.J.; Henry, H.; Dollé, G.; Ritzen, M. J Clin Endocrinol. Metab. 99, 2014, 4328-4335. 24. Eugeni Piller, L.; Besagni, C.; Ksas, B.; Rumeau, D.; Bréhélin, C.; Glauser, G.; Kessler, F.; Havaux, M. Proc. Natl. Acad. Sci. U S A. 2011, 108, 14354-14359. 25. Glauser, G.; Marti, G.; Villard, N.; Doyen, G.A.; Wolfender, J.L.; Turlings, T.C.; Erb, M. Plant J. 2011, 68, 901-911.

18 ACS Paragon Plus Environment

Page 19 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

26. Kaufmann, A.; Walker, S. Rapid Commun. Mass Spectrom. 2013, 30, 347-356. 27. Croley, T.R.; White, K.D.; Callahan, J.H.; Musser, S.M. J. Am. Soc. Mass Spectrom. 2012, 23,1569-1578. 28. Rochat, B.; Kottelat, E.; McMullen, J. Bioanalysis 2012, 4, 2939-2958. 29. Rochat, B.; Peduzzi, D.; McMullen, J.; Favre, A.; Kottelat, E.; Favrat, B.; Tissot, J.D.; Angelillo-Scherrer, A.; Bromirski, M.; Waldvogel, S. Bioanalysis 2013, 5, 2509-2520. 30. Cepurnieks, G.; Rjabova, J.; Zacs, D.; Bartkevics, V. J. Pharm. Biomed. Anal. 2015, 102, 184-192. 31. Sundström, M.; Pelander, A.; Angerer, V.; Hutter, M.; Kneisel, S.; Ojanperä, I. Anal. Bioanal. Chem. 2013, 405, 8463-8474. 32. Bruce, S.J.; Rochat, B.; Béguin, A.; Pesse, B.; Guessous, I.; Boulat, O.; Henry, H. Rapid Commun. Mass Spectrom. 2013, 27, 200-206. 33. Gikas, E.; Bazoti, F.N.; Katsimardou, M.; Anagnostopoulos, D.; Papanikolaou, K.; Inglezos, I.; Skoutelis, A.; Daikos, G.L.; Tsarbopoulos, A. J. Pharm. Biomed. Anal. 2013, 83, 228-236. 34. De Clercq, N.; Julie, V.B.; Croubels, S.; Delahaut, P.; Vanhaecke, L. J. Chromatogr. A. 2013, 1301, 111-121. 35. Kaufmann, A.; Butcher, P. Rapid Commun. Mass Spectrom. 2006, 20, 3566-3572. 36. Vergeynst, L.; Van Langenhove, H.; Joos, P.; Demeestere, K. Anal. Chim. Acta. 2013, 789, 74-82. 37. Van der Heeft, E.; Bolck, Y.J.; Beumer, B.; Nijrolder, A.W.; Stolker, A.A.; Nielen, M.W. J. Am. Soc. Mass Spectrom. 2009, 20, 451-446. 38. Xia, Y.Q.; Lau, J.; Olah, T.; Jemal, M. Rapid Commun. Mass Spectrom. 2011, 25, 2863-2878. Websites 101. FDA; Homepage, Guidance for Industry: Bioanalytical Method Validation, 2013 http://www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm368107.pdf 102. EMA; Homepage, Guideline on bioanalytical method validation 2011 http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2011/08/WC500109686.pdf

19 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 26

Table of Content / Abstract Graphic.

For Table of Contents Only

20 ACS Paragon Plus Environment

Page 21 of 26

Analytical Chemistry

Figure 1

A

LC-HRMS Peak Area of chlorogenic acid in [%] relatively to the area at MEW = ± 28 ppm

LC-HRMS Peak Area (arbitrary units)

250 140

104

200 MEW

150

353.0508 353.0876

100

70

7 14 28

100

98

3

50

0 353.0

1.5

353.1

m/z

0.15

0

0

50

100

150

MEW [ppm]

B

Number of additional LC-HRMS peaks to nilotinib at the analyte RT ± 2 min

in addition to nilotinib

5

Number of LC-HRMS peaks

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

4

nilotinib

interfering ions

MA < 2 ppm

MA > 15 ppm

time

3

8.5

time 8.5

10.5

50

10.5

100

2

20

1

0

15 5 10 5

25 50 75 ACS ParagonMEW Plus Environment [ppm]

100

Analytical Chemistry

Page 22 of 26

Figure 2 6

A

B MAprecision

4

Mass Accuracy [ppm]

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

(SD: 0.9 ppm)

MAhighest (1.0 ppm) 2

MAmean (-0.5 ppm)

MAparameters

MAmean MAprecision MAinterval

MA-related intra-assays I. overtime IIa. calibration IIb. Biomatrix IIc. interscan

0 selection of the greatest MA-parameters

-2

-4

MAinterval (3.2 ppm)

MAlowest (-2.2 ppm)

3 MA-parameters

(│(MAlowest - MAhighest)│)

-6 MA-related intra-assays

 Time [h]

Overall MA assay

 Biomatrix # [N]

 Scan # [N]  Calibration [ng/mL] ACS Paragon Plus Environment

MAmean MAprecision MAinterval

overall

Page 23 of 26

Analytical Chemistry

Figure 3 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Perform Quantitative LC-HRMS analysis

MEW calculation

False detection evaluation

Determine MA-parameters

I MA-related intra-assays

MAover-time

 MAcalibration II

Graphs

MAbiomatrix

MAmean MAprecision

MAinterval

Graph #1



Peak area vs MEW width

Graph #2



Peak # vs MEW width

MAinter-scan

Select the greatest MA-parameters

Determine MEW width Overall MA assay

MAmean overall MAprecision overall MAinterval overall

Check outlier values with (MAinterval overall) >2 (4xMA precision overall)

showing false detections

Calculate MEWassay, MEWoverall and MEWresolution

Select rational MEW width

with MA-parameters and proposed equations

for a validated HRMS method

ACS Paragon Plus Environment

Analytical Chemistry

Figure 4

Mass Accuracy [ppm]

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Page 24 of 26

MA-related intra-assays System suitability

Method dependent IIa.biomatrix IIb. calibration IIc.inter-scan

I. over-time

6

#1

#2

#3

#5

#4

#6

#8

#7

4 2 0 -2 -4

etravirin; N = 18 N = 29

-6 0

N = 24

8 0

8 0

N = 36 12 0

Time [h]

N = 30

N = 140

N = 60 20 0

20

0

30

Matrix # [N]

ACS Paragon Plus Environment

2.5 - 10,000 ng/mL 1

10

N = 24

100 1,000 10,000 0

Levels [ng/mL]

24

Scan # [N]

Page 25 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Analytical Chemistry

Table 1

Intra–assay # MA [ppm] MAhighest MAlowest MAmean assay MAprecision assay MAinterval assay Intra–assay # MEW [ppm] MEWintra-assay MEWoverall MEWresolution

OT1 * OT2

OT3

OT4

0.7 -2.7 -0.8 0.7 3.4

1.0 -2.2 -0.5 0.9 3.2

1.7 -2.5 -0.7 0.9 4.2

1.0 0.1 0.7 0.2 1.1

OT1

OT2

OT3

OT4

3.6

4.1

4.3

OT5 matrix calib. inter-scan Overall assay [ppm] [ppm] 1.2 1.8 -0.6 -0.3 / 0.0 -0.1 -3.2 -4.0 / 0.5 0.8 -1.2 -1.4 1.4 │MAmean overall│ 0.3 0.3 0.6 0.8 0.9 MAprecision overall 1.2 1.8 2.6 3.7 4.2 MAinterval overall

OT5 matrix calib. inter-scan [ppm] 1.5 1.7 2.0 3.6 4.6 5.6 25.6

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Page 26 of 26

Table 2 orbi-MS#1 orbi-MS#2 Q-TOF#1 overall assay

Q-TOF#2

[ppm]

|MAmean overall|

1.4

7.3

3.3

1.6

MAprecision overall

0.9

0.9

3.0

3.4

MAinterval overall

4.2

2.9

9.3

14.9

MEWoverall

5.6

10.2

12.6

16.5

MEWresolution

25.6

19.3

79.3

66.5

Mass Calibration

internal

external

external

internal

R at analyte m/z

50,000

110,000

15,000

20,000

ACS Paragon Plus Environment