Lipid species annotation at double bond position ... - ACS Publications

Subscriber access provided by ALBRIGHT COLLEGE

Article

Lipid species annotation at double bond position level with custom databases by extension of the MZmine 2 open-source software package Ansgar Korf, Viola Jeck, Robin Schmid, Patrick Olaf Helmer, and Heiko Hayen Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b05493 • Publication Date (Web): 20 Mar 2019 Downloaded from http://pubs.acs.org on March 21, 2019

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 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 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.

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 10 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

Lipid species annotation at double bond position level with custom databases by extension of the MZmine 2 open-source software package Ansgar Korfa, Viola Jecka, Robin Schmida, Patrick O. Helmera, Heiko Hayena* a

Institute of Inorganic and Analytical Chemistry, University of Münster, Corrensstraße 30, 48149 Münster, Germany

* Corresponding author: E-mail: [email protected], Tel.: +49 251 83-3 65 76, Fax: +49 251 83-3 60 13

ABSTRACT: In recent years, proprietary and open-source bioinformatics software tools have been developed for the identification of lipids in complex biological samples based on high-resolution mass spectrometry data. These existent software tools often rely on publicly available lipid databases, such as LIPID MAPS®, which, in some cases, do only contain a limited number of lipid species for a specific lipid class. Other software solutions implement their own lipid species databases, which are often confined regarding implemented lipid classes, such as phospholipids. To address these drawbacks, we provide an extension of the widely used open source metabolomics software MZmine 2, which enables the annotation of detected chromatographic features as lipid species. The extension is designed for straightforward generation of custom database for selected lipid classes. Furthermore, each lipid’s sum formula of the created database can be rapidly modified to search for derivatization products, oxidation products, in-source fragments or adducts. The versatility will be exemplified by a liquid chromatography-high resolution mass spectrometry data set with postcolumn Paternò-Büchi derivatization. The derivatization reaction was performed to pinpoint the double bond positions in diacylglyceryltrimethylhomoserine lipid species in a lipid extract of a green algae (Chlamydomonas reinhardtii) sample. The developed Lipid Search module extension of MZmine 2 supports the identification of lipids as far as double bond position level.

Lipids play a crucial role in cell, tissue and organ physiology.1 These biomolecules possess a broad and complex variety of chemical structures, which is mostly defined by acyl and alkyl chain length, degree of unsaturation, their double bond positions, and their stereochemistry. Especially the position of double bonds has an influence on the chemical, biochemical and biophysical properties of lipids, e. g., as shown by Martinez-Seara and co-workers, revealing that the position of double bonds effect membrane properties.2 Therefore, the structural elucidation of any lipid species should be carried out to the double bond position level, if possible. In order to avoid misunderstandings and misinterpretations, Liebisch et al. have proposed a standardization for the annotation of lipids, which is used in this paper.3 Two main trends have been well established, regarding the analysis techniques and instrumentation applied for lipid identification. On the one hand, high resolution mass spectrometry (HRMS) based direct-infusion approaches, also known as shotgun lipidomics, and on the other hand, liquid chromatography (LC) coupled to HRMS.4 In most studies, lipids are only characterized on the sum formula or the lipid chain level using HRMS alone or in combination with tandem mass spectrometry (MS/MS). The development of bioinformatics tools for the identification of lipids in complex biological samples on those two elucidation levels were often addressed in recent years. Mass spectrometry vendors and research groups have published several different software solutions for the purpose of lipid identification.5–15 Some of these solutions, such as ALEX, LipidXplorer or Greazy, are exclusively designed to analyze data

sets of direct-infusion experiments.7,12,13 These tools in combination with the applied data acquisition technique only allow the annotation by the software supported lipid species on sum formula and lipid chain composition level. For the differentiation of structural isomers, e. g., phosphatidylglycerol and bis(monoacylglycero)phosphate, further techniques such as hydrophilic interaction liquid chromatography (HILIC)-HRMS are mandatory.16 For the identification of lipids in LC-MS and LC-MS/MS data sets, software packages such as LipidSearch™ by Thermo Fisher Scientific or SimLipid® by PREMIER Biosoft have been developed, but are limited to vendor formats.8,14 Open-source and platform-independent LC-MS software are often limited to a small number of lipid classes and species regarding their respective databases, e. g., Lipid Hunter is exclusively designed for the identification of phospholipids. LipidFinder is now also linked to one of the most popular lipid database LIPID MAPS®, but relies on other programs such as XCMS for feature detection.10,11,17,18 Projects such as the R-script LipidMatch have targeted the extension of databases, but like LipidFinder the project relies on other software for feature detection.19 Other tools are designed for special purposes, e. g., LPPTiger for the identification of phospholipid oxidation products.5 During the revision process of the present article, Zhang et al. have presented a platform for windows operating systems to annotate phosphatidylcholine and phosphatidylethanolamine species on double bond position level, which is focused on lipid extracts derivatized by means of photo induced Paternò-Büchi (PB) reaction with acetone.20 Besides all the benefits and drawbacks of all the mentioned software, only the latter is addressing the identification of lipids on double bond position level.

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

The ability to dynamically modify databases allows the rapid annotation of lipid derivatization products, which is mandatory to pinpoint a double bond position using ozone induced dissociation (OzID)21–23 or PB reaction.24,25 Both methods have been recently described for shotgun, LC-MS and even imaging lipidomics data sets. Furthermore, many software databases are limited to a small number of lipid classes and species. Even LIPID MAPS® is limited to three entries of lipid species for the lipid class diacylglyceryltrimethylhomoserine (DGTS), which is a major membrane lipid in the green algae Chlamydomonas reinhardtii (C. reinhardtii).26–28 The now in MZmine 229,30 implemented lipid annotation tool allows the user to create a custom database by modifying any lipid classes sum formulas (available since version 2.34). These databases can be processed against feature lists to rapidly annotate lipids, lipid derivatization products and their respective MS/MS product ions, which serve for localization of the double bond position. Therefore, in this work we present new possibilities for HRMSbased lipid identification for large scale lipidomics studies by extending the widely used open-source metabolomics software MZmine 2 for lipid analysis. Localization of double bond positions in lipid chains The position of double bonds in unsaturated fatty acids in complex lipids is not accessible by applying conventional low energy collision-induced dissociation (CID) tandem MS.31 Therefore, derivatization-based techniques to obtain diagnostic fragments in MS/MS experiments of lipids that do pinpoint to a double bond position have become of great interest in the literature. Blanksby and coworkers have successfully applied OzID, which is based on a gasphase reaction of lipid double bonds with ozone. However, applying OzID requires a modification of the MS instrument itself.2123,32–36

As an alternative approach, a photochemical PB reaction was applied by Ma et al. in electrospray-MS as in-source functionalization.37–41 The PB reaction itself is initiated by UV light activated acetone, which forms an oxetane ring with the double bond of lipid chains. The obtained acetone adducts can be fragmented in CIDMS/MS experiments to obtain diagnostic fragments, which do pinpoint a double bonds’ position. The general reaction pathway of derivatization-based techniques to obtain diagnostic fragments in MS/MS experiments is schematically displayed in Figure 1.

Page 2 of 10

Furthermore, using chromatography prior to MS results in less complex spectra and reduced ion suppression, which can improve the sensitivity especially for lipid species of low-abundance. In addition to the here presented software expansion we will demonstrate the first hyphenation of online PB reaction with an HRMS instrument. The acquired data set of a green algae lipid extract will be used as proof of concept for the MZmine 2 software extension. Experimental section Chemicals Acetone (HPLC grade) and ammonium acetate (NH4Ac) was obtained from Sigma Aldrich. Acetonitrile (ACN), methyl tert-butyl ether (MtBE), isopropyl and methanol (HPLC gradient grade) were obtained from VWR International (Darmstadt, Germany). Water purification was performed using a Milli‐ Q® Academic system (18.2 MΩ cm; 0.22 μm filter; Millipore, Molsheim, France). Sample preparation C. reinhardtii algae (wild type) were provided by the group of Prof. Dr. Hippler (Institute of Plant Biochemistry and Biotechnology, University of Münster, Germany). The lipid extraction and preparation of C. reinhardtii algae samples was conducted based on a protocol reported earlier by Korf et al.28 The preparation included liquid-liquid extraction, based on the protocol developed by Matyash et al.42 with addition of a cell wall disruption using a CEM Focused Microwave (CEM Corporation, Matthews, NC, USA) operating at 90°C and 100 W for 1 min. For analysis, the lipid extract was dried using a gentle nitrogen flow at 40 °C, dissolved in 1 mL MtBE/methanol (3:1) and diluted 20-fold in methanol/isopropyl alcohol/buffer (10 mM NH4Ac, pH 5.75, 5% ACN) (32.5:32.5:35). Instrumentation A 1200 series HPLC system (Agilent Technologies, Santa Clara, CA, USA) modified with a 1200 series capillary pump (Agilent Technologies, Waldbronn, Germany) in combination with an Ascentis® Express C18 column (150 × 0.5 mm, 2.7 μm; Supelco, Bellefonte, USA) was used for lipid separation. An aliquot of 2 μL algae sample extract was injected. A binary gradient was utilized composed of 10 mM NH4Ac buffer (pH 5.75) with 5% ACN (A) and methanol (B). The gradient was set up as follows: 0– 0.1 min, 95% B; 0.1–20 min, from 95 to 100% B, and hold at 100% until 45 min; then back within 0.5 min to 95% B for equilibration. The flow rate was set to 15 µL/min and the column oven was operated at 40 °C. The online PB reaction was performed by a post‐ column derivatization conducted as reported earlier by Jeck et al.25 The post-column eluent flow was combined with a constant acetone flow, using an external syringe pump with a flow rate of 15 μL/min. The PB reaction was carried out in a micro‐ flow reactor based on a deactivated fused‐ silica capillary irradiated by a low‐ pressure mercury lamp (primary emission at 254 nm, model 80–1057‐ 01; BHK, Ontario, Canada).

Figure 1. General reaction and fragmentation scheme for the functionalization and fragmentation of a carbon–carbon double bond to obtain diagnostic fragments for the double bond position. At first, this approach was solely applied to direct-infusion experiments. More recently Bednařík et al. accomplished a PB reaction on tissue for double bond localization in lipids, using matrix-assisted laser desorption ionization (MALDI)-MS imaging.24 The PB reaction was carried out utilizing benzaldehyde as MS imaging compatible reagent instead of acetone to preserve the lateral resolution. Jeck et al. achieved the hyphenation of an online PB reaction with LC-MS by means of post-column derivatization.25 Adding chromatography as a further dimension provides more information about the lipid structures, especially regarding structural isomers.

A Q Exactive plus Orbitrap mass analyzer (Thermo Fisher Scientific, Waltham, MA, USA) was utilized for data acquisition. The instrument was operated in positive mode using a heated electrospray source (HESI II) for ionization. Data acquisition was performed in full scan MS/data‐ dependent‐ MS/MS mode using an m/z range from 450 to 1500 and an isolation window of 1 Da. The selected precursor ions were fragmented using higher energy CTrap dissociation (HCD) with stepped normalized collision energy at 25, 30, and 35 eV. Precursors were dynamically selected using an inclusion list, which contained a database of all possible DGTS lipid species ([M+H]+) and their respective PB products ([M+C3H6O+H]+) in a range of 28 to 44 carbon atoms and 0 to 10 double bonds in both acyl chains combined. The list was created


Page 3 of 10 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


with the here presented extension to MZmine 2 (Figure 3). The resolution was set to 70,000 (full width at half maximum (FWHM) at m/z = 200) for full scan profile spectra and to 35,000 (FWHM at m/z = 200) for centroid MS/MS spectra. The automatic gain control (AGC) target was set to 1e6 with a maximum injection time of 100 ms for full scan spectra and to 1e5 with a maximum injection time of 50 ms for MS/MS spectra. The spray voltage was set to 3.5 kV with a sheath gas (nitrogen) flow rate of 17 a.u. and an auxiliary gas (nitrogen) flow rate of 6 a.u. The capillary temperature was set to 250°C, while the drying gas temperature was set to 50°C. The data acquisition time was set from 1 min to 45 min. Data processing Prior to data processing, raw files were converted to the open format mzML using the MSConvert software, which is part of the cross-platform ProteoWizard software toolkit.44 An optimized MZmine 2 workflow was developed for feature list generation. The workflow included the exact mass detection algorithm, with a noise level set to 2e4. Chromatogram building was performed including a minimum ion time span of 0.1 min, a minimum peak height set to 2e4, and the relative m/z tolerance set to 5 ppm. The next step was chromatogram deconvolution by applying the local minimum search algorithm, including a chromatographic threshold of 90% and a minimum peak top/edge ratio of 2. The latter is calculated by the ratio of the most intense data point (top) and the lowest data point (edge). A minimum absolute intensity was set

to 2e4 and a maximum peak duration range was set to 10 min. Isotopic features were removed by applying an isotopic peaks grouper algorithm including a relative m/z tolerance of 5 ppm, a retention time tolerance of 0.1 min, the monotonic shape parameter set to false, two as the maximum allowed charge, and keeping only the most intense isotope as the final result, which is here the monoisotopic peak. The parameters outlined above are further defined in the MZmine 2 integrated help documentation. Results and Discussion Applying the developed MZmine 2 extension for lipid identification resulted in the putative identification of 44 DGTS lipid species on sum formula level, based on their respected accurate masses, retention times and obtained head group fragments by means of MS/MS experiments. 42 of the 44 lipids were identified on acyl chain level, observing neutral losses of acyl chains in MS/MS experiments, which already revealed the presence of two or sometimes even three structural isomers for 12 of the 42 lipid species. For 6 of the 44 on sum formula level identified lipids an addition of acetone was annotated based on its accurate mass and its coelution with its respective labeled lipid species. By analyzing the MS/MS spectra of each of the PB products, further 15 possible structural isomers which differ in double bond positions were putatively identified. All these results, which are visualized in Fig-

ure 2 in form of a network are 100% software driven. The network Figure 2. Network of annotated DGTS species. DGTS class level (pink); DGTS species sum formula level (blue); DGTS species chain level (green); DGTS species double bond position level (orange).



describes the diversity of the found DGTS species. Starting from the middle, the lipid class (pink) can be extended to several species at sum formula level (blue). The diversity of the species is already visible at the acyl chain level (green), which becomes even more sufficient at the double bond position level (orange). The presented software extension was used for feature annotation on different identification levels under consideration of the Lip-

Page 4 of 10

idomics Standards Initiative guidelines for MS lipid ion monitoring.28 Figure 3 (top) shows a screenshot of the Lipid Search module parameter setup dialog. Here, the user selects the lipid classes of interest (e. g. DGTS), the minimum and maximum number of carbon atoms and summed double bonds in all lipid chains. The lipid species can be further characterized on chain level by selecting the checkbox ”Search for lipid class specific fragments in MS/MS spectra”. The lipid class specific fragments were derived from publicly available databases such as LipidBlast and the literature.45–47

Figure 3. Screenshots of the MZmine 2 Lipid Search module; Top: Parameter setup window; Bottom: Visualization window of DGTS. database.


Page 5 of 10 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


Based on the observed signals in the MS/MS spectra, the software searches for the class specific fragments and reconstructs the lipid structure on chain level and adds information about the head group fragment. Prior to processing, the user can review the generated custom database in form of a table and Kendrick mass plots, displayed in Figure 3 (bottom). Kendrick mass plots have recently been presented in the literature as a tool for the graphical identification of lipids.28 In Kendrick mass plots, homologous compounds appear on a horizontal line with respect to the chain length (Figure 3, bottom, left) and the number of double bonds (Figure 3, bottom, right), depending on the selected Kendrick mass base. The table includes information on the lipid class, sum formula, abbreviation, ion type, exact mass and MS/MS fragments. Furthermore, a color coded status regarding the possible interference with other lipid species present in the database is visualized in a further column and in the Kendrick mass plots. To pinpoint the double bond position of a lipid species, PB reaction was performed as post-column derivatization. The observed acetone adducts of the lipid species have been annotated using the “Search for lipid modification” parameter of the Lipid Search module. Every generated lipid species can be modified by adding or subtracting sum formulas. The as PB product annotated features have been further analyzed by searching for diagnostic fragments in the MS/MS spectra of the selected PB product precursor. Therefore, we have developed a new MS/MS visualizer module that displays all MS/MS spectra of the selected feature in a single frame. The annotation of diagnostic fragments can also be performed directly in the selected MS/MS spectra with the Lipid Search module. For this, the "Search for lipid modification" parameter was used again to generate all possible diagnostic fragments. Signals identified as diagnostic fragments are labeled directly in the spectra. All entered modifications can be exported to a csv file, which can be imported for another analysis. Therefore, we have provided a csv file of all possible PB-products, which can easily be applied on other datasets using the import function of the “Search for lipid modification” parameter. The results for DGTS 32:1 are exemplified in Figure 4. Figure 4 (top left) shows the species’ extracted ion chromatogram (EIC) with its coeluting PB product below and the corresponding MS/MS spectra on the right. The MS/MS spectra contains the two characteristic head group fragments of DGTS spe-

cies (C7H14NO2 and C10H22NO5) and two neutral losses of fatty acids (FA) (A = [M-FA(16:1)+H]+ and B = [M-FA(16:0)+H]+) which allow the identification on acyl chain level. The corresponding MS/MS spectra of the PB product contain two diagnostic fragments ([M-C12H24+H]+ and [M-C9H18O+H]+), which do pinpoint the double bond position Δ7 of FA(16:1). All possible resulting double bond positions observed for all identified lipid species are listed in Table 1 and are visualized in Figure 2. The results are in agreement with previous studies. Giroud et. al have characterized lipids in C. reinhardtii, revealing the overall fatty acid composition of DGTS species.27 The interpretation of the diagnostic fragments of all DGTS species revealed the double bond positions Δ7 and Δ9 in C16:1 chains, Δ9 and Δ11 in C18:1, (Δ9, Δ12) in C18:2, (Δ5, Δ9, Δ12) and (Δ9, Δ12, Δ15) in C18:3, and (Δ5, Δ9, Δ12, Δ15) in C18:4. Please note that the shorthand nomenclature was used with ΔX notation of double bond position, in which the carbon number is counted from the ester group.3

Figure 4. EIC of DGTS 32:1 [M+H]+ (top, left) and corresponding MS/MS spectra (top, right). The MS/MS spectra shows both characteristic head group fragments for DGTS species and the neutral losses A = [M-FA(16:1)+H]+ and B = [M-FA(16:0)+H]+. EIC of DGTS 32:1 PB product [M+C3H6O+H]+ (bottom, left) and corresponding MS/MS spectra (bottom, right). The MS/MS spectra contains two diagnostic fragments which do pinpoint the double bond position Δ7.

Table 1: Results of annotated DGTS species on different identification level (sum formula, chain and double bond). Accurat e Mass m/z

δm/m Intensity Retentio [ppm (height of n ] chromatogram time ) [min]

682.560 1

2.17

257915

14.2

678.528 8 696.575 9 710.591 1

2.20

81679

8.7

2.06

250392

17.7

2.50

15958474

21.8

708.575 7

2.20

1929408

15.5

Sum fomula level

Chain level

PBproduct precursor m/z /

DGTS(30:1 FA(16:1)_FA(14:0 ) ) FA(16:0)_FA(14:1 ) DGTS(30:3 / / ) DGTS(31:1 FA(16:1)_FA(15:0 / ) ) DGTS(32:1 FA(16:0)_FA(16:1 768.637 ) ) 3 DGTS(32:2 FA(16:0)_FA(16:2 ) ) FA(18:2)_FA(14:0 )

/

Double bond position level

/

/ / FA(16:0)_FA(16:1(Δ7)) FA(16:0)_FA(16:1(Δ9)) /



706.560 8

1.14

1002684

12.9

704.544 6 702.529 4 702.529 4 724.606 7

1.96

1795028

11.3

1.34

302683

8.3

1.34

67409

8.8

2.64

890393

25.7

722.591 4

2.03

515697

19.3

720.575 4 718.560 3 738.622 1 736.606 9

2.67

816430

15.6

1.81

98198

12.29

2.80

6118886

29.8

2.35

26366236

23.9

734.590 9

2.66

47401016

19.3

732.575 2

2.87

7171529

15.1

732.575 2

2.87

4734055

13.7

Page 6 of 10

DGTS(32:3 FA(18:3)_FA(14:0 / ) ) FA(16:0)_FA(16:3 ) DGTS(32:4 FA(16:0)_FA(16:4 / ) ) DGTS(32:5 FA(16:1)_FA(16:4 / ) ) DGTS(32:5 FA(16:1)_FA(16:4 / ) ) DGTS(33:1 FA(18:1)_FA(15:0 / ) ) FA(17:1)_FA(16:0 ) DGTS(33:2 FA(18:2)_FA(15:0 / ) ) FA(17:2)_FA(16:0 ) DGTS(33:3 FA(18:3)_FA(15:0 / ) ) DGTS(33:4 FA(18:4)_FA(15:0 / ) ) DGTS(34:1 FA(18:1)_FA(16:0 / ) ) DGTS(34:2 FA(18:1)_FA(16:1 794.649 ) ) 4 FA(18:2)_FA(16:0 )

DGTS(34:3 FA(18:3)_FA(16:0 792.633 ) ) 4 FA(18:2)_FA(16:1 ) FA(18:1)_FA(16:2 ) DGTS(34:4 FA(18:3)_FA(16:1 / ) ) FA(18:4)_FA(16:0 ) DGTS(34:4 FA(18:3)_FA(16:1 / ) ) FA(18:1)_FA(16:3 )

730.560 2

1.94

937352

10.8

DGTS(34:5 )

728.544 4 726.529 1 752.638 1 752.638 1 750.622 3 750.622 3

2.23

313436

9.07

1.63

504145

7.5

2.44

153943

32.2

2.44

108556

33.6

2.51

363200

28.1

2.51

144931

26.4

DGTS(34:6 ) DGTS(34:7 ) DGTS(35:1 ) DGTS(35:1 ) DGTS(35:2 ) DGTS(35:2 )

FA(18:2)_FA(16:2 ) FA(18:3)_FA(16:2 ) FA(18:4)_FA(16:1 ) FA(18:3)_FA(16:3 ) FA(18:3)_FA(16:4 ) FA(19:1)_FA(16:0 ) FA(18:1)_FA(17:0 ) FA(18:2)_FA(17:0 ) FA(19:1)_FA(16:1 ) FA(18:1)_FA(17:1 )

/

/ / / /

/

/ / / FA(16:0)_FA(18:2(Δ9,Δ12)) FA(16:1(Δ9))_FA(18:1(Δ11)) FA(16:1(Δ9))_FA(18:1(Δ9)) FA(16:1(Δ7))_FA(18:1(Δ11)) FA(16:1(Δ7))_FA(18:1(Δ9)) FA(16:0)_FA(18:3(Δ5,Δ9,Δ12)) FA(16:0)_FA(18:3(Δ9,Δ12,Δ15)) FA(18:2(Δ9,Δ12))_FA(16:1(Δ9)) FA(18:1(Δ9))_FA(16:2(Δ7,Δ10))

/

/

/

/

/

/

/

/

/

/

/

/

/

/

/

/


Page 7 of 10 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 748.606 6 748.606 6 746.591 2 746.591 2 766.653 7 764.637 9 764.637 9 762.622 4 762.622 4 760.606 9 758.590 9

2.63

507305

23.9

2.63

153279

20.6

2.30

242267

16.4

2.30

108375

15.4

2.37

140488

37.2

2.56

1118037

30.5

2.56

575360

32.3

2.39

3218572

24.8

2.39

1259265

28.1

2.19

5817434

20.3

2.66

11396514

15.1

758.590 9 756.575 4 754.559 8 752.543 3 774.624 5 790.653 0 788.638 7

2.66

1922785

23.9

2.54

10254537

11.9

2.37

1684016

9.8

3.62

118642

7.8

-0.32

52515

23.2

3.14

88354

32.6

1.48

123100

28.9

DGTS(35:3 ) DGTS(35:3 ) DGTS(35:4 ) DGTS(35:4 ) DGTS(36:1 ) DGTS(36:2 ) DGTS(36:2 ) DGTS(36:3 ) DGTS(36:3 ) DGTS(36:4 ) DGTS(36:5 )

FA(18:3)_FA(17:0 / ) FA(18:2)_FA(17:1 / ) FA(18:3)_FA(17:1 / ) FA(18:3)_FA(17:1 / ) FA(18:0)_FA(18:1 / ) FA(18:0)_FA(18:2 / ) FA(18:0)_FA(18:2 / ) FA(18:1)_FA(18:2 / ) FA(18:0)_FA(18:3 / ) FA(18:1)_FA(18:3 818.649 ) 7 FA(18:2)_FA(18:3 816.632 ) 9 FA(18:1)_FA(18:4 ) FA(18:2)_FA(18:3 / ) FA(18:2)_FA(18:4 814.917 ) 4 FA(18:3)_FA(18:4 / ) / /

DGTS(36:5 ) DGTS(36:6 ) DGTS(36:7 ) DGTS(36:8 ) DGTS(37:4 FA(19:1)_FA(18:3 ) ) DGTS(38:3 FA(20:1)_FA(18:2 ) ) DGTS(38:4 FA(20:1)_FA(18:3 ) )

/ / / / / / / / / FA(18:1(Δ11))_FA(18:3(Δ5,Δ9,Δ12)) FA(18:2(Δ9,Δ12))_FA(18:3(Δ5,Δ9,Δ12)) FA(18:2(Δ9,Δ12))_FA(18:3(Δ9,Δ12,Δ15))

/ FA(18:2(Δ9,Δ12))_FA(18:4(Δ5,Δ9,Δ12,Δ1 5)) / /

/

/

/

/

/

/

and the source code will be made available online in the GitHub repository (https://github.com/mzmine/mzmine2).

Conclusion Using the developed MZmine 2 software extension for lipid identification of a green algae extract of C. reinhardtii, acquired by means of online PB reaction coupled to LC-HRMS, resulted in the identification of 44 DGTS membrane lipids on sum formula level and 55 lipids on chain level. Furthermore, 15 possible isomers have been putatively annotated on double bond position level based on their determined double bond positions and the recombination of acyl chains. The presented software solution allowed the analysis of the complex data set in less than a day, including data processing steps, ranging from raw data conversion to feature list generation. The here presented results are the first successful application of an online PB reaction, by means of hyphenation of LC-HRMS. Furthermore, due to the high potential of database customization using the lipid modification parameter, the software extension can be used for other tasks, such as lipid adduct annotation or identification of other lipid derivatization products, including different lipid classes, such as phospholipids, making the software not limited to the in this work used PB-reaction. This enables further structural elucidation possibilities of lipids using other techniques, such as OzID. Using MZmine 2 in combination with the here presented extension can also be used for large scale studies due to a batch mode integration. The extension is available since MZmine version 2.34

Associated content A supporting information document is included, which contains information on the utilized inclusion list and a workflow diagram of the developed software. Furthermore, a user manual for the developed Lipid Search module is available in a separate document, and a .txt file “PB_Fragments_Modification_List.txt” is included, which can be used with the Lipid Search module. Acknowledgment The authors like to thank Prof. Dr. Hippler (Institute of Plant Biochemistry and Biotechnology, University of Münster, Germany) for providing C. reinhardtii algae samples. Financial support by the Deutsche Forschungsgemeinschaft (DFG, Bonn Germany) (Grant No. HA 4319/9‐ 1 and INST 211/802-1) is also gratefully acknowledged. Conflicts of interest The authors declare no competing financial interest.



References (1) Wenk, M. R. The emerging field of lipidomics. Nat. Rev. Drug Discov. 2005, 4, 594–610. (2) Martinez-Seara, H.; Róg, T.; Pasenkiewicz-Gierula, M.; Vattulainen, I.; Karttunen, M.; Reigada, R. Interplay of unsaturated phospholipids and cholesterol in membranes: Effect of the double-bond position. Biophys. J. 2008, 95, 3295–3305.

Page 8 of 10

(13) Kochen, M. A.; Chambers, M. C.; Holman, J. D.; Nesvizhskii, A. I.; Weintraub, S. T.; Belisle, J. T.; Islam, M. N.; Griss, J.; Tabb, D. L. Greazy: Open-source software for automated phospholipid tandem mass spectrometry identification. Anal. Chem. 2016, 88, 5733–5741. (14) Isaac, G.; Mcdonald, S.; Astarita, G. Automated lipid identification using UPLC / HDMS E in combination with SimLipid. Waters - Application note 2011, 1–8.

(3) Liebisch, G.; Vizcaíno, J. A.; Köfeler, H.; Trötzmüller, M.; Griffiths, W. J.; Schmitz, G.; et al. Shorthand notation for lipid structures derived from mass spectrometry. J. Lipid Res. 2013, jlrM033506.

(15) Hartler, J.; Trötzmüller, M.; Chitraju, C.; Spener, F.; Köfeler, H. C.; Thallinger, G. G. Lipid Data Analyzer: Unattended identification and quantitation of lipids in LC-MS data. Bioinformatics 2011, 27, 572–577.

(4) Schwudke, D.; Schuhmann, K.; Herzog, R.; Bornstein, S. R.; Shevchenko, A. Shotgun lipidomics on high resolution mass spectrometers. Cold Spring Harb. Perspect. Biol. 2011, 3, 1–13.

(16) Vosse, C.; Wienken, C.; Cadenas, C.; Hayen, H. Separation and identification of phospholipids by hydrophilic interaction liquid chromatography coupled to tandem high resolution mass spectrometry with focus on isomeric phosphytidylglycerol and bis(monoacylglycero)phosphate. J. Chromatogr. A 2018, 1565, 105–113.

(5) Ni, Z.; Angelidou, G.; Hoffmann, R.; Fedorova, M. LPPtiger software for lipidome-specific prediction and identification of oxidized phospholipids from LC-MS datasets. Sci. Rep. 2017, 7, 15138. (6) Goracci, L.; Tortorella, S.; Tiberi, P.; Pellegrino, R. M.; Di Veroli, A.; Valeri, A.; Cruciani, G. Lipostar, a comprehensive platform-neutral cheminformatics tool for lipidomics. Anal. Chem. 2017, 89, 6257–6264. (7) Herzog, R.; Schuhmann, K.; Schwudke, D.; Sampaio, J. L.; Bornstein, S. R.; Schroeder, M.; Shevchenko, A. Lipidxplorer: A software for consensual cross-platform lipidomics. PLoS One 2012, 7, e29851. (8) Peake, D.; Yokoi, Y.; Wang, J. A new lipid software workflow for processing Orbitrap-based global lipidomics data in translational and systems biology research. Thermo Scientific Application note 2013, 1–7. (9) Ni, Z.; Angelidou, G.; Lange, M.; Hoffmann, R.; Fedorova, M. LipidHunter identifies phospholipids by high-throughput processing of LC-MS and shotgun lipidomics datasets. Anal. Chem. 2017, 89, 8800–8807. (10) O’Connor, A.; Brasher, C. J.; Slatter, D. A.; Meckelmann, S. W.; Hawksworth, J. I.; Allen, S. M.; O’Donnell, V. B. LipidFinder: A computational workflow for discovery of lipids identifies eicosanoid-phosphoinositides in platelets. JCI Insight 2017, 2, e91634. (11) Fahy, E.; Alvarez-Jarreta, J.; Brasher, C. J.; Nguyen, A.; Hawksworth, J. I.; Rodrigues, P.; Meckelmann, S.; Allen, S. M.; O’Donnell, V. B.; Valencia, A. LipidFinder on LIPID MAPS: Peak filtering, MS searching and statistical analysis for lipidomics. Bioinformatics 2018, 1–3. (12) Husen, P.; Tarasov, K.; Katafiasz, M.; Sokol, E.; Vogt, J.; Baumgart, J.; Nitsch, R.; Ekroos, K.; Ejsing, C. S. Analysis of lipid experiments (ALEX): A software framework for analysis of high-resolution shotgun lipidomics data. PLoS One 2013, 8, e79736.

(17) Schmelzer, K.; Fahy, E.; Subramaniam, S.; Dennis, E. A. The Lipid Maps initiative in lipidomics. Methods Enzymol. 2007, 432, 171–183. (18) Smith, C. A.; Want, E. J.; O’Maille, G.; Abagyan, R.; Siuzdak, G. XCMS: Processing mass spectrometry data for metabolite profiling using nonlinear peak alignment, matching, and identification. Anal. Chem. 2006, 78, 779–787. (19) Koelmel, J. P.; Kroeger, N. M.; Ulmer, C. Z.; Bowden, J. A.; Patterson, R. E.; Cochran,; et al. LipidMatch: an automated workflow for rule-based lipid identification using untargeted high-resolution tandem mass spectrometry data. BMC bioinformatics, 2017, 18, 331. (20) Zhang, W.; Zhang, D.; Chen, Q.; Wu, J.; Ouyang, Z.; Xia, Y. Online photochemical derivatization enables comprehensive mass spectrometric analysis of unsaturated phospholipid isomers. Nat. Commun., 2019, 10, 79. (21) Thomas, M. C.; Mitchell, T. W.; Harman, D. G.; Deeley, J. M.; Nealon, J. R.; Blanksby, S. J. Ozone-induced dissociation: Elucidation of double bond position within mass-selected lipid ions. Anal. Chem. 2008, 80, 303–311. (22) Poad, B. L. J.; Pham, H. T.; Thomas, M. C.; Nealon, J. R.; Campbell, J. L.; Mitchell, T. W.; Blanksby, S. J. Ozone-induced dissociation on a modified tandem linear ion-trap: Observations of different reactivity for isomeric lipids. J. Am. Soc. Mass Spectrom. 2010, 21, 1989–1999. (23) Mitchell, T. W.; Pham, H.; Thomas, M. C.; Blanksby, S. J. Identification of double bond position in lipids: From GC to OzID. J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 2009, 877, 2722–2735. (24) Bednařík, A.; Bölsker, S.; Soltwisch, J.; Dreisewerd, K. On-tissue Paternò-Büchi reaction for localization of carbon-carbon double bonds in phospho- and glycolipids by matrix-assisted

ACS Paragon Plus Environment

8

Page 9 of 10 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


laser desorption ionization mass spectrometry imaging. Angew. Chemie Int. Ed. 2018, 57, 12092–12096. (25) Jeck, V.; Korf, A.; Vosse, C.; Hayen, H. Localization of double bond positions in lipids by tandem mass spectrometry succeeding high performance liquid chromatography with post-column derivatization. Rapid Commun. Mass Spectrom. 2018, DOI: https://doi.org/10.1002/rcm.8262.

(36) Thomas, M. C.; Mitchell, T. W.; Harman, D. G.; Deeley, J. M.; Murphy, R. C.; Blanksby, S. J. Elucidation of double bond position in unsaturated lipids by ozone electrospray ionization mass spectrometry. Anal. Chem. 2007, 79, 5013–5022. (37) Ma, X.; Xia, Y. Pinpointing double bonds in lipids by Paternò-Büchi reactions and mass spectrometry. Angew. Chemie Int. Ed. 2014, 53, 2592–2596.

(26) Sud, M.; Fahy, E.; Cotter, D.; Brown, A.; Dennis, E. A.; Glass, C. K.; Merrill, A. H.; Murphy, R. C.; Raetz, C. R. H.; Russell, D. W.; et al. LMSD: LIPID MAPS structure database. Nucleic Acids Res. 2007, 35, D527-32.

(38) Ma, X.; Zhao, X.; Li, J.; Zhang, W.; Cheng, J.-X.; Ouyang, Z.; Xia, Y. Photochemical tagging for quantitation of unsaturated fatty acids by mass spectrometry. Anal. Chem. 2016, 88, 8931–8935.

(27) Giroud, C.; Gerber, A.; Eichenberger, W. Lipids of Chlamydomonas reinhardtii. Analysis of molecular species and intracellular site(s) of biosynthesis. Plant Cell Physiol. 1988, 29, 587– 595.

(39) Ma, X.; Chong, L.; Tian, R.; Shi, R.; Hu, T. Y.; Ouyang, Z.; Xia, Y. Identification and quantitation of lipid C=C location isomers: A shotgun lipidomics approach enabled by photochemical reaction. Proc. Natl. Acad. Sci. 2016, 113, 2573–2578.

(28) Korf, A.; Vosse, C.; Schmid, R.; Helmer, P. O.; Jeck, V.; Hayen, H. Three-dimensional Kendrick mass plots as a tool for graphical lipid identification. Rapid Commun. Mass Spectrom. 2018, 32, 981–991.

(40) Ren, J.; Franklin, E. T.; Xia, Y. Uncovering structural diversity of unsaturated fatty acyls in cholesteryl esters via photochemical reaction and tandem mass spectrometry. J. Am. Soc. Mass Spectrom. 2017, 28, 1432–1441.

(29) Katajamaa, M.; Miettinen, J.; Orešič, M. MZmine: Toolbox for processing and visualization of mass spectrometry based molecular profile data. Bioinformatics 2006, 22, 634–636.

(41) Zhao, Y.; Zhao, H.; Zhao, X.; Jia, J.; Ma, Q.; Zhang, S.; Zhang, X.; Chiba, H.; Hui, S.-P.; Ma, X. Identification and quantitation of C═C location isomers of unsaturated fatty acids by epoxidation reaction and tandem mass spectrometry. Anal. Chem. 2017, 89, 10270–10278.

(30) Pluskal, T.; Castillo, S.; Villar-Briones, A.; Orešič, M. MZmine 2: Modular framework for processing, visualizing, and analyzing mass spectrometry-based molecular profile data. BMC Bioinformatics 2010, 11, 395. (31) Hancock, S. E.; Poad, B. L.; Batarseh, A.; Abbott, S. K.; & Mitchell, T. W. Advances and unresolved challenges in the structural characterization of isomeric lipids. Anal. Biochem. 2017, 524, 45-55. (32) Marshall, D. L.; Pham, H. T.; Bhujel, M.; Chin, J. S. R.; Yew, J. Y.; Mori, K.; Mitchell, T. W.; Blanksby, S. J. Sequential collision- and ozone-induced dissociation enables Assignment of relative acyl chain position in triacylglycerols. Anal. Chem. 2016, 88, 2685–2692. (33) Kozlowski, R. L.; Mitchell, T. W.; Blanksby, S. J. A rapid ambient ionization-mass spectrometry approach to monitoring the relative abundance of isomeric glycerophospholipids. Sci. Rep. 2015, 5. (34) Pham, H. T.; Maccarone, A. T.; Thomas, M. C.; Campbell, J. L.; Mitchell, T. W.; Blanksby, S. J. Structural characterization of glycerophospholipids by combinations of ozone- and collisioninduced dissociation mass spectrometry: The next step towards “top-down” lipidomics. Analyst 2014, 139, 204–214. (35) Poad, B. L. J.; Green, M. R.; Kirk, J. M.; Tomczyk, N.; Mitchell, T. W.; Blanksby, S. J. high-pressure ozone-induced dissociation for lipid structure elucidation on fast chromatographic timescales. Anal. Chem. 2017, 89, 4223–4229.

(42) Matyash, V.; Liebisch, G.; Kurzchalia, T. V; Shevchenko, A.; Schwudke, D. Lipid extraction by methyl-tert-butyl ether for high-throughput lipidomics. J. Lipid Res. 2008, 49, 1137–1146. (43) Liebisch, G.; Ekroos, K.; Hermansson, M.; Ejsing, C. S. Reporting of lipidomics data should be standardized. Biochimica et Biophysica Acta-Molecular and Cell Biology of Lipids 2017, 1862, 747-751. (44) Chambers, M. C.; Maclean, B.; Burke, R.; Amodei, D.; Ruderman, D. L.; Neumann, S.; Gatto, L.; Fischer, B.; Pratt, B.; Egertson, J.; et al. A cross-platform toolkit for mass spectrometry and proteomics. Nat. Biotechnol. 2012, 30, 918–920. (45) Kind, T.; Liu, K.-H.; Yup Lee, D.; DeFelice, B.; Meissen, J. K.; Fiehn, O. LipidBlast - in-silico tandem mass spectrometry database for lipid identification. Nat. Methods 2013, 10, 755–758. (46) Melo, T.; Alves, E.; Azevedo, V.; Martins, A. S.; Neves, B.; Domingues, P.; Calado, R.; Abreu, M. H.; Domingues, M. R. Lipidomics as a new approach for the bioprospecting of marine macroalgae - unraveling the polar lipid and fatty acid composition of Chondrus crispus. Algal Res. 2015, 8, 181–191. (47) Lu, N.; Wei, D.; Chen, F.; Yang, S. T. Lipidomic profiling and discovery of lipid biomarkers in snow alga Chlamydomonas nivalis under salt stress. Eur. J. Lipid Sci. Technol. 2012, 114, 253–265.


9



Page 10 of 10

10

Lipid species annotation at double bond position ... - ACS Publications

Recommend Documents