High-Resolution, Accurate Mass Measurements and Metabolite Identification: An Automated Approach Using Fragment Prediction in Combination with Fragment Ion Search (FISh)Janine Wank1, Winfried Wagner-Redeker1, Anthony Taylor2, Rose Herbold3, Paul-Gerhard Lassahn1 1Swiss BioAnalytics AG, Basel, Switzerland; 2Thermo Fisher Scientific, Reinach, Switzerland; 3Thermo Fisher Scientific, San Jose, CA, USA

Application Note: 537

Key Words

• LTQOrbitrapSeries

• MassFrontier

• FragmentIonSearch(FISh)

• MetaboliteIdentification

• StructuralElucidation

IntroductionStructural elucidation of putative metabolites via LC/MS is a time consuming process of manual interpretation of MSn data and requires a strong background in metabolism pathways as well as gas-phase fragmentation pathways. To evaluate an automated software workflow for metabolite identification and structural elucidation, we processed high-resolution, accurate mass data using the Fragment Ion Search (FISh)™ processing tool in Thermo Scientific Mass Frontier 7.0 software. Data-dependent MSn data was used for component detection and for the identification of metabolites based on theoretically calculated fragments of the parent structure.

GoalTo evaluate an automated workflow for metabolic profiling and structural elucidation by using Fragment Ion Search (FISh) processing in Mass Frontier™ 7.0 software.

Experimental Conditions

SamplePreparationThe investigated metabolites were derived from ticlopidine, a potent thienopyridine antiplatelet drug known to be extensively metabolized. Ticlopidine was incubated with human liver S9 enzymes. The final experimental concen-trations in the 1 mL incubation mixture were 100 mM KPO4 (pH 7.4), 3 mM NADPH, 3.8 μM ticlopidine, and S9 human liver fraction (BD Bioscience). After the addition of S9 human liver fraction, the incubation mixtures were homogenized gently and placed into a water bath (37 °C). Aliquots of the reaction solution were withdrawn after 0, 30, and 60 minutes, and the reaction was quenched by the addition of acetonitrile. The individual samples were centrifuged and analyzed by LC/MS.

LiquidChromatographySeparations were performed using a Thermo Scientific Hypersil GOLD column (50 x 2.1 mm, 1.9 μm particle size) with a Thermo Scientific Accela 600 pump and a Thermo Scientific Open Accela autosampler. The chromatographic conditions were as follows:

Mobile phase A: Water with 0.1% formic acid

Mobile phase B: Acetonitrile with 0.1% formic acid

Flow rate: 150 µL/min

Gradient: Time (min) %A %B

0.0 90 10

2.0 90 10

16.0 50 50

18.0 5 95

23.0 5 95

23.1 90 10

28.0 90 10

MassSpectrometryMS analysis was carried out with a Thermo Scientific LTQ Orbitrap XL hybrid mass spectrometer in positive ion mode using one high-resolution full MS scan followed by three high-resolution MS2 scans. Precursor selection was done in the data-dependent operation mode where the most intense ion of the previous scan was selected for fragmentation. The MS conditions were as follows:

MS resolution: 60,000 FWHM at m/z 400

MS2 resolution: 7,500 FWHM at m/z 400

Full MS mass range (m/z): 150 - 800

Collision Energy: CID: 35%

In additional experiments, an inclusion list containing the calculated exact masses of putative metabolites was added to the method (Figure 1).

Data AnalysisRaw data was processed using Mass Frontier 7.0 software, including the Chromatogram Processor and Database Manager modules. The Fragments and Mechanisms, FISh, FISh Explanation, and Compare Spectra processing tools were used. Multiple Mass Defect Filtering was completed by using Thermo Scientific MetWorks 1.3 software.

Results and DiscussionThe structures of ticlopidine and its known metabolites1 are highlighted in Figure 2. The extracted ion chromato-grams of these metabolites (0 min versus 30 min S9 liver fraction incubation; mass accuracy 5 ppm) are displayed and annotated in Figure 3.

The high resolution accurate mass data (data-

dependent MS/MS mode) was directly exported to the Mass Frontier 7.0 software program. After the automated removal of noise and baseline signals, the theoretical

calculation of possible fragments of ticlopidine was performed. (See Figure 1 for the schematic workflow and Figure 4 for the annotation of all major peaks in the MS/MS spectra of ticlopidine.) Next, a general list of possible Phase I biotransformations was applied, and the LC/MS chromatogram was processed by a component detection algorithm. FISh was then used to screen the detected components and the corresponding spectral ion trees to identify putative metabolites of ticlopidine.

The workflow and FISh data processing results are summarized and illustrated in Figure 5. After the determination of the theoretical fragments of ticlopidine and specification of the Phase l

modifications, component detection and FISh processing were used to screen for putative metabolites. The list of predicted fragments was automatically extended by the mass shifts of the modifications through the definition of the expected biotransformations of ticlopidine. The list and the combined ion chromatograms of putative metabolites are displayed after this process. FISh provided an explanation of the MS/MS data including the structures of the fragments and the possible modifications (highlighted in red in Figure 5) for each detected component. For the metabolites M1 and M5, the modifications were further confirmed by a comparison between the observed and theoretical fragments of the putative structures of the metabolites. All fragment ions were explained.

General Data-Dependent MS/MS

(high-resolution, accurate mass measurements)Data-Dependent MS/MS with parent mass list

(high-resolution, accurate mass measurements)

Additional: Mass-defect-filtering (MetWorks)

Metabolite Identification (Mass Frontier):

• Theoretical fragmentation of parent

• Noise and baseline removal

• Component detection

• FISh (Fragment Ion Search)

List of putative metabolites

Experimental planning and sample preparation

Calculated, exact masses of

possible metabolites (MetWorks)

Calculated, exact masses of

possible metabolites (MetWorks)

Figure 1. Schematic workflow of the automated data processing for metabolite identification

N

SCl

Ticlopidine, [C14H14NSCl+H]+m/z 264.0608

N

ClS

N

C l

O

S Oor isom ers N

SCl

O

N+

SCl

N+

SCl

N

SCl

O

N

SClO H

M1, Hydroxyticlopidine,[C14H14NOSCl+H]+, m/z 280.0557

M2, 2-Oxoticlopidine,[C14H14NOSCl+H]+, m/z 280.0557

M3, Dihydropyridinium Metabolite,[C14H13NSCl]+, m/z 262.0452

M4, Thienopyridinium Metabolite,[C14H11NSCl]+, m/z 260.0295

M5, Ticlopidine N-oxide,[C14H14NOSCl]+, m/z 280.0557

M6, Ticlopidine S-oxide dimer (TSOD),[C28H28N2O2S2Cl2+H]+, m/z 559.1042

Figure 2. Ticlopidine and its known metabolites

RT: 1.0 - 14.0 SM: 5G

2 4 6 8 10 12 14

Time (min)

0

1

2

3

4

5

6

7

8

9

Re

lative

Ab

un

da

nce

0

1

2

3

4

5

6

7

8

9

Re

lative

Ab

und

an

ce

8.3

6.2

11.5

2.48.2

6.57.8

10.0

NL: 6.09E6m/z =260.0282-260.0308+262.0439-262.0465+264.0595-264.0621+280.0543-280.0571+559.1014-559.1070 F: FTMS + p ESIFull ms [150.00-800.00] MS13052011_11004_05

NL: 2.81E6m/z =260.0282-260.0308+262.0439-262.0465+264.0595-264.0621+280.0543-280.0571+559.1014-559.1070 F: FTMS + p ESIFull ms [150.00-800.00] MS13052011_11004_07

Ticlopidine

M6M3

M5

M1*

M1/M2**M4

Incubation time: 0 min

Incubation time: 30 min

* Fragment ions at m/z 262 indicating the loss of H2O which is most probably for M1.

**Fragment ions at m/z 262 less abundant than for peak at RT 2.4 min; the presence of an isomer

of M1 cannot be excluded.

Figure 3. Extracted ion chromatograms of ticlopidine and its known metabolites M1-M6 (mass accuracy 5 ppm, retention-time 1.0-14.0 min)

RT: 0 .00 - 14 .00 S M : 7G

0 2 4 6 8 10 12 14

Tim e (m in)

0

10

20

30

40

50

60

70

80

90

100

Re

lative

Ab

un

da

nce

0

10

20

30

40

50

60

70

80

90

100

Re

lative

Ab

un

da

nce

9 .00

8 .056.17

11 .292.41 9.62

8.98

6.207 .87

7.44

8.222.389.71

NL: 1.65E6m/z =260.0282-260.0308+262.0439-262.0465+264.0595-264.0621+280.0543-280.0571+559.1014-559.1070 F: FTMS + p ESIFull ms [150.00-800.00] MS13052011_11004_18

NL: 1 .16E 5

m /z= 125.0147-125 .0159 F : F TM S + c E S I d F ull m s2 M S 13052011_11004_18

Extracted ion chromatograms of

the fragment m/z 125.0153

Ticlopidine

M6

M4

M5

M1 M1/M2

Extracted ion chromatograms of

M1-M6

m/z 125.0153

M3

Figure 4. Extracted ion chromatograms of the ticlopidine metabolites M1-M6 and of the fragment m/z 125.0153 (C7H6Cl, mass accuracy 5 ppm, incubation time 30 min)

Extracted ion chromatograms of

detected related components

M4

M5M1

M1/M2

M3

M6*

Load LC/MSn raw data file

into Mass Frontier 7.0

Generate theoretical

fragments of ticlopidine

Define modifications

Start Component Detection

and FISh processing

M1

M1

M2

M4

M3M4

M5M6*

*in-source fragment of the

S-oxide dimer M6

FISh Explanation for M1 FISh Explanation for M5

Extracted ion chromatograms of

detected related components

M4

M5M1

M1/M2

M3

M6*

n raw data file

ontier 7.0

Generate theoretical

ffragments of ticlopid nine

Define modifications

onent Detectioon

h processing

M1

M1

M2

M4

M3M4

M5M6*

*in-source fragment of the

S-oxide dimer M6

FISh Explanation for M1 FISh Explanation for M5

Annotation of the MS/MS spectrum of M1 Annotation of the MS/MS spectrum of M5

with predicted fragmentswith predicted fragments

m/z125.0153

m/z138.0372m/z154.0418

m/z

m/z138.0372

m/z111.0263

125.0153

Figure 5. Explanation of the automated approach for metabolite identification using fragment prediction in combination with Fragment Ion Search (FISh)

262.05

m/z

262.045

262.026

262.045

262.026

RT:0.00 - 14.98 SM: 5G

0 2 4 6 8 10 12 14

Time (min)

0

10

20

30

40

50

60

70

80

90

100

Rela

tive A

bundance

0

10

20

30

40

50

60

70

80

90

100

Rela

tive A

bundance

RT: 8.12

RT: 8.05

7.56

NL: 1.04E5

m/z =262.0439-262.0465 F:

FTMS + p ESI Full ms

[150.00-800.00] MS

Genesis

13052011_11004_18

NL: 1.32E5

m/z =260.0282-260.0308 F:

FTMS + p ESI Full ms

[150.00-800.00] MS

13052011_11004_18

262.05

m/z

262.045

262.026

262.045

262.026

Extracted ion-chromatogram of M3

(m/z 262.0452)

Extracted ion-chromatogram of M4

(m/z 260.0295)

M4 (35Cl)

M3 (37Cl)

M3 (35Cl)

Simulated isotopic pattern for

M3/M4 (ratio 4/1)

Measured isotopic pattern for

M3/M4 (ratio 4/1, single spectrum)

M4

(37Cl)

M3

(35Cl)

260 261 262 263 264 265

m/z

0

10

20

30

40

50

60

70

80

90

100

0

10

20

30

40

50

60

70

80

90

100

Rela

tive A

bundance

262.045

264.041260.029

263.047261.031 265.044

260.148

262.045

264.042

260.030

263.048

265.045261.033

NL:8.77E4

13052011_11004_18#477 RT: 8.19 AV: 1 F: FTMS + p ESI Full ms [150.00-800.00]

NL:1.44E6

C14H11NSCl*22.00 + C14H13NSCl*100.00: p(gss, s/p:40) Chrg1R: 85000 Res.Pwr. @FWHM

Figure 6. Co-elution of the metabolites M3/M4 and the benefit of high-resolution, accurate mass data

Table 1. Summary of the Fragment Ion Search (FISh) processing results

Expected FIShwithmass Metabolite FISh defectfiltereddata

M1

M2

M3

M4

M5

*M6*

* M6 was detected through the identification of an in-source fragment

The results of the FISh processing and a comparison between expected and detected metabolites1 are displayed in Table 1. All known metabolites, including the unusual S-oxide dimer metabolite M6, were detected and identified by the automated FISh processing approach of the original raw data file. The unusual S-oxide dimer metabolite M6 was detected by FISh through the identification of an in-source fragment (Figure 6).

An additional comparison was performed to assess if the automated FISh processing workflow could be com-bined with the Multiple Mass Defect Filtering (MMDF) tool within MetWorksTM 1.3 software. MMDF is a popu-lar data processing tool used in metabolite identification to remove endogenous background ions.2 The same FISh processing workflow was repeated on a MMDF prepro-cessed raw data file. The results showed no differences in the detected metabolites in comparison to without MMDF preprocessing (Table 1). However, the FISh processing time for the MMDF preprocessed file was shorter. This demon-strates that MMDF and FISh can be effectively combined for fast and automated metabolite identification without incurring false negatives.

To demonstrate the benefit of high-resolution, accurate mass data, the measured and simulated Full MS data of the co-eluting metabolites M3 and M4 are displayed in Figure 6. The high-resolution, accurate mass measurement of the isotopic pattern provided verification of the putative identification of the two metabolites.

Part of Thermo Fisher Scientific

www.thermofisher.comLegal Notices: ©2016 Thermo Fisher Scientific Inc. All rights reserved. Mass Frontier and Fragment Ion Search (FISh) are trademarks of HighChem, Ltd. All other trademarks are the property of Thermo Fisher Scientific and its subsidiaries. This information is presented as an example of the capabilities of Thermo Fisher Scien-tific Inc. products. It is not intended to encourage use of these products in any manners that might infringe the intellectual property rights of others. Specifications, terms and pricing are subject to change. Not all products are available in all countries. Please consult your local sales representative for details. AN63435_E 08/16S

In addition to these

offices, Thermo Fisher

Scientific maintains

a network of represen

tative organizations

throughout the world.

Africa-Other +27 11 570 1840Australia +61 3 9757 4300Austria +43 1 333 50 34 0Belgium +32 53 73 42 41Canada +1 800 530 8447China +86 10 8419 3588Denmark +45 70 23 62 60 Europe-Other +43 1 333 50 34 0Finland/Norway/ Sweden +46 8 556 468 00France +33 1 60 92 48 00Germany +49 6103 408 1014India +91 22 6742 9434Italy +39 02 950 591Japan +81 45 453 9100LatinAmerica +1 561 688 8700MiddleEast +43 1 333 50 34 0Netherlands +31 76 579 55 55NewZealand +64 9 980 6700Russia/CIS +43 1 333 50 34 0SouthAfrica +27 11 570 1840Spain +34 914 845 965Switzerland +41 61 716 77 00UK +44 1442 233555USA +1 800 532 4752

ConclusionMetabolic profiling and structural elucidation of possible metabolites via LC/MS/MS was simplified and accelerated by Fragment Ion Search (FISh) processing in Mass Frontier 7.0 software. This workflow can be combined with Multiple Mass Defect Filtering (MMDF) in MetWorks 1.3 software for reduced processing times without incurring false negatives. The availability of high-resolution, accurate mass LC/MS/MS data minimizes the possibility of false positives in the generated list of related components. All known metabolites of ticlopidine in a S9 human liver fraction incubation were detected and identified using this automated processing workflow, including the unusual S-oxide dimer metabolite. This is interesting to notebecause FISh was able to provide an indication for thepresence of this metabolite just through the identificationof an in-source fragment.

References1. Talakad, J.C.; Shah, M.B.; Walker, G. S.; Xiang, C.; Halpert, J.R.;

Dalvie, D. Comparison of In Vitro Metabolism of Ticlopidine by Human Cytochrome P450 2B6 and Rabbit Cytochrome P450 2B4 Drug Metab. Dispos. 2011, 39(3), 539-550.

2. Huang, Y.; Liu, S.; Miao, S.; Jeanville, P.M. Using Multiple Mass Defect Filters and Higher Energy Collisional Dissociation on an LTQ Orbitrap XL for Fast, Sensitive and Accurate Metabolite ID; Thermo Fisher Scientific Application Note 417, 2008.