Mass spectrometry based lipid(ome) analyzer and molecular platform: a new software to interpret and...

Research article

Received: 10 June 2012 Revised: 23 December 2012 Accepted: 30 December 2012 Published online in Wiley Online Library

(wileyonlinelibrary.com) DOI 10.1002/jms.3163

Mass spectrometry based lipid(ome) analyzerand molecular platform: a new software tointerpret and analyze electrospray and/ormatrix-assisted laser desorption/ionizationmass spectrometric data of lipids: a case studyfrom Mycobacterium tuberculosisVaratharajan Sabareesha,b*† and Gurpreet Singhc{

Mass Spectrometry based Lipid(ome) Analyzer andMolecular Platform (MS-LAMP) is a new software capable of aiding in interpret-ing electrospray ionization (ESI) and/or matrix-assisted laser desorption/ionization (MALDI) mass spectrometric data of lipids.
The graphical user interface (GUI) of this standalone programme is built using Perl::Tk. Two databases have been developedand constituted within MS-LAMP, on the basis of Mycobacterium tuberculosis (M. tb) lipid database (www.mrl.colostate.edu) andthat of Lipid Metabolites and Pathways Strategy Consortium (LIPID MAPS; www.lipidmaps.org).
Different types of queries entered through GUI would interrogate with a chosen database. The queries can be molecular mass(es) or mass-to-charge (m/z) value(s) and molecular formula. LIPID MAPS identifier also can be used to search but not for M. tblipids. Multiple choices have been provided to select diverse ion types and lipids. Satisfying to input parameters, a glimpse ofvarious lipid categories and their population distribution can be viewed in the output. Additionally, molecular structures of lipidsin the output can be seen using ChemSketch (www.acdlabs.com), which has been linked to the programme. Furthermore, aversion of MS-LAMP for use in Linux operating system is separately available, wherein PyMOL can be used to view molecularstructures that result as output from General Lipidome MS-LAMP.

The utility of this software is demonstrated using ESI mass spectrometric data of lipid extracts of M. tb grown under twodifferent pH (5.5 and 7.0) conditions. Copyright © 2013 John Wiley & Sons, Ltd.

Supporting information may be found in the online version of this article.

Keywords: software; ESI; MALDI; Mycobacterium tuberculosis; lipids; LIPID MAPS; Perl:Tk

* Correspondence to: Varatharajan Sabareesh, Proteomics and StructuralBiology Unit, CSIR-Institute of Genomics and Integrative Biology (IGIB). E-mail:[email protected]/[email protected]

† Current address: Centre for Bio-Separation Technology, Vellore Institute of Tech-nology University, Vellore, Tamilnadu 632014, India. E-mail: [email protected].

{ Current address: Department of Chemical & Biomolecular Engineering, NationalUniversity of Singapore, Singapore 117576.

a Proteomics and Structural Biology Unit, Council of Scientific and IndustrialResearch, Institute of Genomics and Integrative Biology (CSIR-IGIB), Mall Road,New Delhi, 110007, India

b CSIR-IGIB Extension Centre (Naraina), A93/94, Naraina Industrial Area - Phase I,Naraina, New Delhi, 110028, India

c Department of Biotechnology, School of Biotechnology Engineering, LovelyProfessional University, Jalandhar-Delhi G.T. Road (NH-1), Phagwara, Punjab,144402, India

465

Introduction

Amongst the ‘omics’ kind of investigations, during the lastdecade, ‘lipidomics’ emerged and now has matured into adiscipline, which has almost become routine in different kindsof biological, biomedical and healthcare research. A key reasonfor such a rapid growth can be attributed to innovations in massspectrometry (MS), which is perhaps currently the mainstay ofthis area of research, primarily involving electrospray ionization(ESI) and/or matrix-assisted laser desorption and ionization(MALDI) mass spectrometers.[1–3] Several lipids have been identi-fied and investigated predominantly by electron impact orelectron ionization as part of gas chromatography–MS, and theseinclude mostly non-polar and volatile lipids, for instance, fattyacids and esters of fatty acids.[2,4,5] Fast atom bombardment–MS enabled characterization of polar, non-volatile lipids such asglycerophospholipids and sphingolipids.[2,4,6] Further, the adventof ESI and MALDI, in conjunction with sophisticated massanalyzers (e.g. ion cyclotron resonance and orbitrap), enabled -omic type of approach to characterize lipids. Simultaneousimprovements in liquid chromatography (LC) as well have con-tributed for the current level of development.[2] ESI and MALDI

J. Mass Spectrom. 2013, 48, 465–477

mass spectrometric methods not only aid in identifying lipids ata large scale but also are useful for quantification purposes.[7–11]

Huge data are recorded be it shotgun approach (direct infusion)

Copyright © 2013 John Wiley & Sons, Ltd.

http://www.acdlabs.com

V. Sabareesh and G. Singh

466

or by LC coupled MS (LC–MS). To process and enable analyses ofsuch vast sets of data computational tools are therefore required.The ultimate aim is to assign signals or peaks of mass-to-charge(m/z) values in mass spectra to one or more lipid(s) or lipid class(es), through which identification of lipids is accomplished.Amidst already available several software tools that have been

designed to facilitate analysis and interpretation of lipidomicmass spectrometric data,[12–22] Mass Spectrometry based Lipid(ome) Analyzer and Molecular Platform (MS-LAMP) is a newaddition. MS-LAMP possesses several unique features, althougha few aspects are common between MS-LAMP and earlierreported softwares. The main objective of any software in suchcontexts is to facilitate in ascribing as many signals (m/z values)as possible in a mass spectrum to one or more lipid(s), eventuallyaiding in identifying lipids in a given sample. MS-LAMP is suitedto assign mass spectral peaks obtained from ESI and MALDI.Despite the availability of both Web-based and standalone

versions of various softwares, efforts towards design of MS-LAMPwere initiated so as to have a tool customized specifically to caterto in-house requirements involving characterization of lipidome.Further impetus to create MS-LAMP originated particularly duringthe study of lipids from Mycobacterium tuberculosis (M. tb). Thelipid database of M. tb, ‘Mtb LipidDB’ had just been uploadedby the Mycobacteria Research Laboratories of Colorado StateUniversity (www.mrl.colostate.edu),[23] when experiments werebegun in our laboratory to investigate changes in the lipidprofiles that could occur on growing M. tb at two different pHconditions. A brief survey of this database showed that therewere 2518 lipids, among which 2512 were found to be unique.The lipids in this database were classified following the defini-tions proposed by the Lipid Metabolites and Pathways StrategyConsortium (LIPID MAPS, available at www.lipidmaps.org).[24–26]

Thus, it was decided to include the database of M. tb as well asthat from LIPID MAPS (which consists of 37 056 lipid molecularstructures, as of June 2012) in MS-LAMP. Accordingly, the graph-ical user interface (GUI) of MS-LAMP was designed. The programcan be downloaded from http://ms-lamp.igib.res.in.

Materials and methods

Growth of Mycobacterium tuberculosis and extraction of lipids

A laboratory strain of M. tb was grown in two different pHconditions, 5.5 and 7.0, employing two distinct 7H9 media. Fromthe pelleted cells, hexane (hexane only) was used for extractionof cell wall lipids, whereas total lipid extraction was achieved bychloroform (CHCl3)/methanol (CH3OH), 2 : 1 (v/v)).

Mass spectrometry

The mass spectrometric experiments were conducted on an LTQOrbitrap XL (Thermo Scientific) using conventional ESI realizedby Ion Max source. All data acquisitions were carried out innegative ion mode only. The lipid extracts were directly infusedinto the mass spectrometer by a syringe pump. The flow ratewas varied in the range 10–20 ml/min. The spray voltage wasmaintained at about �6 kV. The other source parameter settingswere as follows: sheath gas flow, 45 arbitrary units; capillaryvoltage, �15 V; capillary temperature, 300 �C; and tube lens,�100 V. Tandem mass spectrometric (MS/MS) experiments werecarried out following collision induced dissociation (CID) usinghelium as collision gas. CID was accomplished in LTQ, whereas

wileyonlinelibrary.com/journal/jms Copyright © 2013 Jo

the product ion analyses were carried out by Orbitrap. The para-meters for acquiring MS/MS data shown here are as follows:precursor ion isolation width, m/z 3 for precursor ion m/z1136.18 and m/z 1 for m/z 1164.20; activation q, 0.25; activationtime, 30ms; and collision energy, 34. Data processing was carriedout with Qual Browser version 2.0.7 SP1 (Thermo Scientific).

Making of MS-LAMP

On the basis of the information of lipids available in databasesof the LIPID MAPS consortium (www.lipidmaps.org) and theMycobacteria Research Laboratories, Colorado State University(www.mrl.colostate.edu), two databases have been constructedand incorporated within MS-LAMP. The details about lipids suchas molecular formula, molecular structures, etc. as present in thesetwo databases, were downloaded in the form of comma separatedvalues (.csv) file format. Subsequently, these were converted into .xls files, using which different kinds of analyses were carried out.From the insights obtained through these database analyses(see Results section), the algorithm for this software was planned,and accordingly, modifications were carried out to these databases.

Thus, the two databases are called here as ‘Mycobacteriumtuberculosis (M. tb) Lipidome MS-LAMP’ and ‘General LipidomeMS-LAMP’. There are 2518 lipids in M. tb Lipidome MS-LAMP,whereas General Lipidome MS-LAMP consists of 37 572 lipids.Among the 37572, molecular mass is known for 37 056 lipids;however, molecular mass information is not available for 516 lipidsin the LIPID MAPS database (see Supporting information file:Lipids_No-Mol-mass_General-MS-LAMP_updated_June-2012.xls).

Back-end programming of MS-LAMP was carried out usingStrawberry Perl (5.12.3). Information about lipid(s) had been separatedinto different files. Upon submitting a query, for faster data acquisitionand rapid retrieval of information about lipid(s), hash (data structures)is created dynamically making the runtime faster. The GUI for thestandalone programme was developed using Perl::Tk module.Furthermore, Microsoft SAPI module has been utilized to providespeech-based introduction to the Windows version of this software.

Molecular structures for lipid entries in General LipidomeMS-LAMP are available as MDLmol files obtained from LIPIDMAPS,whereas structures for M. tb Lipidome MS-LAMP are available as .jpeg image files, adopted from Mtb LipidDB. ChemSketch(Advanced Chemistry Development Inc./Labs: www.acdlabs.com/resources/freeware/chemsketch/) has been linked to MS-LAMP,and hence, the molecular structures can be viewed using MDLmol files. However, ChemSketch is handy for General LipidomeMS-LAMP only. In case of M. tb Lipidome MS-LAMP, the structuresof lipids need to be viewed by the use of .jpeg files, as MDmol filesare absent in the current version of Mtb LipidDB.

The results of the search can be saved in three different fileformats: (1) comma-separated values (.csv); (2) text file (.txt);and (3) Microsoft Excel (.xls). Spreadsheet-WriteExcel module avail-able from Cpan was used to write to a cross-platform Excel binaryfile. On saving the results as .xls file, the first worksheet shows thesummary of the result as seen in the ‘Results of the Query’ window.The second and third worksheets respectively contain a bar graphand pie chart representation of the results, which provides a quicklook at the results. Further, each worksheet is devoted to a lipidcategory/group/subgroup, showing lipids respective to thatcategory/group/subgroup that have been yielded for a query.When the results are saved as .txt file, the first few lines in this fileprovide a glimpse of the results, as shown in the first worksheetof the .xls file, i.e. as depicted in the ’Result of the Query’ window.

hn Wiley & Sons, Ltd. J. Mass Spectrom. 2013, 48, 465–477

http://www.mrl.colostate.edu

http://www.lipidmaps.org

http://ms-lamp.igib.res.in


http://www.acdlabs.com/resources/freeware/chemsketch/

http://www.acdlabs.com/resources/freeware/chemsketch/

MS-LAMP: software for lipids’ mass spectrometry

The source code of the program will be provided upon request.New databases can be appended, and the program can be used;however, the newly added database needs to be modifiedaccording to the functional make-up of MS-LAMP. The program(as .exe file) is freely available for download from the Webpagehttp://ms-lamp.igib.res.in. MS-LAMP is also compatible with Linuxand has been tested on Ubuntu 11.04. Notes on installationof the same are available within the package (as .rar file), which isalso freely available for download from the same webpage, asmentioned earlier.

Results

Description of MS-LAMP

MS-LAMP is a standalone programme consisting of GUI, throughwhich queries can be entered to seek lipids. Utilizing the recentlyreported lipid database of M. tb (www.mrl.colostate.edu) and

Figure 1. Snapshots of different windows of MS-LAMP: (a) main window of M(M. tb) Lipidome MS-LAMP and General Lipidome MS-LAMP; (b) window of Mthis is obtained on clicking button (1) of the main window (see (a)).

J. Mass Spectrom. 2013, 48, 465–477 Copyright © 2013 John W

that of LIPID MAPS (www.lipidmaps.org), two separate databasescorresponding to M. tb and LIPID MAPS have been built forMS-LAMP. We call these databases as M. tb Lipidome MS-LAMPand General Lipidome MS-LAMP (Fig. 1(a)). For a queried input,the programme searches by interrogating with a chosen databaseand then provides the output. Matching to the queried input, theoutput contains a list of lipids. To facilitate viewing of molecularstructures of lipids resulting in the output, ChemSketch (www.acdlabs.com) application has been linked to MS-LAMP.

Input details

Figure 1(a) shows the first GUI window of MS-LAMP containingclickable buttons (indicated as (1) and (2)) to open a programcorresponding to a database, which opens in a new window.Figure 1(b) shows a snapshot of GUI of M. tb Lipidome MS-LAMPthat opens as a new window on clicking the button (1) depictedin the first GUI. Four options of query types are displayed: (1) single

S-LAMP containing buttons corresponding to Mycobacterium tuberculosis. tb Lipidome MS-LAMP containing buttons of four different query types;

iley & Sons, Ltd. wileyonlinelibrary.com/journal/jms

467

http://ms-lamp.igib.res.in






468

molecular mass or m/z value; (2) list of molecular masses orm/z values; (3) molecular mass or m/z range; and (4) molecularformula. For a search, only one of the aforementioned queries isto be opted.Button (2) of the first window (Fig. 1(a)) corresponds to the

opening of the GUI of General Lipidome of MS-LAMP, whichcontains five query types (Fig. S1, Supporting information). Inaddition to the previously described four query types, anotheroption to search lipids has been included only in GeneralLipidome MS-LAMP, which is query type 5, where LM ID (LIPIDMAPS ID as designated by the LIPID MAPS consortium) canbe used to search lipids within General Lipidome MS-LAMP.An LM ID is a unique 12-character identifier consisting ofalphabets and arabic numerals defined by the LIPID MAPSconsortium for easier and systematic tracking of lipids; forexample, LMSP01040007 denotes a sphingophospholipid, whosecommon name is C17 sphingosine-1-phosphocholine, whereasLMSP01040008 represents C16 sphingosine. The query type 5 isnot made available for searching M. tb lipidome MS-LAMPbecause only some lipids (as of February 2012, 57 lipids)have been designated with the LM IDs in Mtb LipidDB (www.mrl.colostate.edu).Every query type opens in a new GUI window. The GUI

windows of both query types 1 and 2 contain the followingoptions: (i) a type of ion or molecular mass; (ii) window range;and (iii) lipid category(ies).

(i) Type of ion: Both negative and positive ions have beenincluded. The six different types of negative ions are [M�H]�,[M+HOAc�H]�, [M+Cl]�, [M+Li� 2H]�, [M+Na� 2H]�

and [M+K� 2H]�. The eight positively charged ions are[M+H]+, [M+NH4]

+, [M+Li]+, [M+Na]+, [M+K]+, [M+2Li�H]+,[M+2Na�H]+ and [M+2K�H]+. In total, the GUI displayscheck buttons for these 14 ions and for molecular mass (Fig. 2(a)). A query will consist of any one of these 14 ions or merelymolecular mass.

(ii) Window range: Depending on the resolution of the massspectrometer used and subjective to the user’s requirementof accuracy, this will be useful. Any one of the values of

Figure 2. Snapshot of ‘query type 1’ window from M. tb Lipidome MS-LAMlecular mass: queried m/z, 1115; ion type, [M+NH4]

+; window range, +/�0.2


0.005, 0.01, 0.05, 0.1, 0.25, 0.5, 0.75 and 1.0 can be selected(Fig. 2).

(iii) Lipid categories: As already mentioned, MS-LAMP utilizes thedatabase of the LIPID MAPS consortium, from which theGeneral Lipidome MS-LAMP database is made. In doing so,the eight-category classification as defined by the LIPID MAPSconsortium is adopted by MS-LAMP. The eight categories arefatty acyls (FA), glycerolipids (GL), glycerophospholipids (GP),prenol lipids (PR), polyketides (PK), saccharolipids (SL),sphingolipids (SP) and sterol lipids (ST). A search may bedirected to all these eight categories together; else, one ormore category(ies) could be selected. In addition to theclassification of eight categories, three more groups wereformed, which are ‘all phospholipids’, ‘halogenated lipids’and ‘sulfur-containing lipids’. ‘All phospholipids’ group containssubgroups of ‘sphingophospholipids’ and ‘other phosphory-lated lipids’. Because there is already a category for GP in LIPIDMAPS, it may be noted that in ‘all phospholipids’ group, thereare preferences allocated to only ‘sphingophospholipids’ and‘other phosphorylated lipids’. However, all phospholipids groupcovers phosphorylated lipids present in all categories, meaningincluding GP.

To look for a particular type of GL, the user can explicitly pointthe query to ‘monoglycerides’, ‘diglycerides’, ‘triglycerides’ and‘other glycerolipids’ (other GLs). These groups and subgroups havebeen made without deviating from the classification characteristicsof LIPID MAPS. By having such specific groups and subgroups,computationally the search process gets simplified. Moreover, italso enables a user to target the queries in a very specific manner.Thus, altogether, there are 18 preferences (i.e. 18 check buttons)to search for lipids (Fig. 2). Amongst these 18, one or more lipidcategories can be chosen.

Furthermore, search can also be directed to a ‘main class’ or‘subclass’ within a lipid category. To do this, it is essential tounderstand the codes defined to address categories, main classand subclass, as delineated in LIPID MAPS. In this regard, a help file(in .pdf format) has been given that can be obtained on clicking the‘Need Help?’ button provided beside this option. From this help file,

P, showing an example of searching lipids on querying single m/z or mo-5, and search directed to all eight categories.





appropriate codes can be known to focus the search to a particularclass within a category. This search feature is available in querytypes 1–4.

Query type 1: Fig. 2 shows screenshot of the GUI of querytype 1 of M. tb Lipidome MS-LAMP, illustrating an example ofsearching lipids for the following chosen parameters: queriedm/z, 1115; ion type, [M+NH4]

+; window range, +/�0.25; andsearch directed to ‘all eight categories’. On submitting this query,the user will be directed to the Result of the Query windowdepicting a brief overview of population distribution of differentlipid categories matching to this input (Fig. 3).

Query type 2: Fig. S2 (Supporting information) illustrates anexample of identifying lipids by submitting a list of m/z valuesusing query type 2. These m/z values are experimentallydetermined from ESI mass spectrometric investigation of hexaneextract of a laboratory strain of M. tb grown in a medium main-tained at pH 7.0 (Fig. 7 and Table 6). With this query type, twodifferent kinds of files can be uploaded: (i) list of m/z and nointensity values and (ii) list of m/z and relative intensity values.When a file containing relative intensity values is uploaded, theuser can set a threshold value for relative intensities. Only thosem/z values whose intensities are above the set threshold wouldbe queried in, and the corresponding lipids can be identified.Further, the user-defined threshold option will be useful indistinguishing signals due to contribution of isotopes from themolecular ion signals. When there are two peaks, whose m/zvalues differ by 1 mass unit, it is essential to determine whetherthe peak at higher m/z value is a contribution from isotope(s)(e.g. 13C) or due to another lipid, whose molecular mass is 1 massunit higher. For instance, in the ESI mass spectrum of an FA lipid ofmolecular formula C89 H174 O3, the intensities of the molecularion and the first, second and third isotope peaks would be in theratio 100 : 98 : 48 : 16, respectively, whereas those of a GL, whosemolecular formula is C71 H132 O6, would be 100 : 78 : 32 : 9,respectively. Thus, in such cases, the user-defined threshold func-tionality can be of immense utility; the user can vary this threshold,analyze the results and interpret in accordance with the submitted

Figure 3. Snapshot of ‘Result of the Query’ window from M. tb Lipidome MS‘query type 1’, as shown in Fig. 2.


list of m/z and intensity values (see section on Regarding IsotopePeaks for further discussion on this aspect).

Query type 3: The query type 3 is for searching lipids bymeans of defined molecular mass or m/z range. This can becombined with choices of only (i) a type of ion or molecular massand (iii) lipid category(ies). The option of (ii) window range is notincluded in its GUI. Fig. S3 (Supporting information) depicts acase from query type 3 of General Lipidome MS-LAMP, whereasFig. S4 (Supporting information) depicts the steps involved inviewing molecular structures of lipids (that yield as output shownin Fig. S3b (Supporting information)) in ChemSketch. The stepsillustrated to view molecular structure of lipids in Fig. S4(Supporting information) are the same to visualize molecularstructures in the case of query types 1, 2 and 4.

Query type 4: When searching lipids using molecular formula(query type 4), only (iii) lipid categories needs to be selected(Fig. S5, Supporting information).

Query type 5: Information on lipids can be sought using LMID as well, using query type 5 of General Lipidome MS-LAMP;an example is shown in Fig. S6 (Supporting information).

Output details

For the query types 1–4, the output window titled as Result of theQuery displays a list of different lipid categories (as queried by theuser) and the number of lipids in each of those categories. Fromsuch an output representation, one gets a quick look at the diversetype of lipids matching to the queried criteria. In .xls output file, thefirst worksheet shows the summary of the input and output details.The second and third worksheets illustrate the results in the form ofbar graph and pie chart, respectively. Such representations presenta pictorial glance of the population distribution of lipids in variouscategories as queried in. Thereafter, each worksheet is devoted toeach lipid category or group; in a worksheet, a list of lipids belong-ing to a lipid category or group is shown. The results of the querycan be saved in any one of three different file formats as per user’schoice, .csv, .txt or .xls.

-LAMP, depicting results corresponding to the parameters submitted into


469


470

Figure 4 depicts the flowchart summarizing the functioning ofMS-LAMP.

Databases analyses

Mycobacterium tuberculosis Lipidome MS-LAMP

On the basis of Mtb LipidDB (www.mrl.colostate.edu), thisdatabase has been constructed. There are 2518 lipids that havebeen classified into eight categories following the classificationscheme of LIPID MAPS. The molecular masses of these lipids aredistributed in the range 302.24571 – 3064.965305Da (Table 1).Strikingly, there are no SP and ST in this database (Fig. 5,Table S1a, Supporting information) suggesting that eithersuch lipids are yet to be discovered in this parasite or perhapsM. tb lacks the machinery for biosynthesis of these two lipidcategories. The FA category predominantly comprises mycolates(537 mycolates out of 751; Table S1b, Supporting information).In addition to 1295 GP, only two PR and five PK are phosphory-lated (Fig. 5, Table S1b, Supporting information). Amongst164 sulfolipids, 163 belong to the SL category, and the other isa PR lipid (Fig. 5, Table S1b, Supporting information). Interest-ingly, 41% of this database is composed of ‘glycerophosphoinosi-tolglycans (GPIG)’ (Fig. 5), and thus far, no halogenated lipidshave been found from M. tb.

Figure 4. Flowchart providing an overview of functioning of MS-LAMP.

Table 1. Total number of lipids in Mycobacterium tuberculosis (M. tb)Lipidome MS-LAMP and their molecular mass range*

Molecular mass range (in Da) 302.24571 – 3064.965305

Total entries 2518

Unique entries (based onmolecular mass) 2506

* Source: Mycobacteria Research Laboratories, Colorado StateUniversity (www.mrl.colostate.edu) (as of September 2012).


General Lipidome MS-LAMP

This database, which has been built on the basis of LIPIDMAPS (last updated, June 2012), consists of 37 572 lipidsclassified into eight categories. Molecular mass is known for37 056 lipids, whereas there is no information on molecularformula (and hence no molecular mass) for 516 lipids (Table 2and Supporting information file: Lipids_No-Mol-mass_General-MS-LAMP_updated_June-2012.xls). The categories are furtherdivided into classes and subclasses. The molecular massesof lipids in this database are in the range 46.00548–4722.081313Da (Table 3). The population distribution of lipidsacross these eight categories is summarized in Fig. 6 (also seeTable S2a, Supporting information). Detailed assessment oflipids in this database revealed some intriguing features.

Besides 8005 GP, there is non-negligible population of phos-phorous-containing (or phosphorylated) lipids in other categoriesas well. Barring GP, when considering other seven categoriescollectively, there are 527 phosphorylated lipids, notable being89 phosphorylated FA and 350 sphingophospholipids (Fig. 6and Table S2a, Supporting information). Further, the databasecomprises 458 sulfur-containing lipids (Fig. 6 and Table S2a,Supporting information). Interestingly, there is only one haloge-nated sphingolipid, and there are no halogenated GL, GP and SLin a total of 345 (Fig. 6 and Tables S2a, Supporting information).Altogether, there are 118 fluorine-containing lipids, 93 lipidshave chlorine, 112 are brominated and 22 possess iodine in theirmolecular structures (Table S2b, Supporting information).

Introduction of new lipid groups

These analyses impelled to create a new classification format forMS-LAMP that comprised ‘groups’, namely Halogenated Lipids,All Phospholipids and Sulfur-containing lipids (Figs 5 and 6;Tables S1 and S2, Supporting information). All Phospholipidsgroup consists of subgroups, ‘sphingophospholipids’ and ‘Otherphosphorylated lipids’. GP is not a subgroup of All Phospholipids,although a search directed to this group would include hits fromGP as well. It is apparent from the survey that the populations oflipids in these groups are not insignificant, particularly in GeneralLipidome MS-LAMP (Figs 5 and 6; Tables S1 and S2). Further, theformation of these groups and subgroups is still based on chem-ical properties, which are distinct and yet without diverging fromthe eight-category scheme of LIPID MAPS. It may be recalled thatthe eight-category sorting of LIPID MAPS was chemically basedand driven by distinct hydrophobic and hydrophilic elements.[24]

Thus, in addition to the eight categories, introduction of thesegroups enables to focus the search in a more specific manner,which will be useful for targeted and comparative investigations.

Regarding isotope peaks

It is well known that the abundance of second and third isotopesof a given molecular ion exceeds the intensity of the molecularion, when the number of carbon (C) atoms in a molecule is morethan about 90 or so and when the molecule possesses morehalogens, in particular Cl and Br. Table 4 shows the number oflipids having more than 90 C atoms in each of the eight catego-ries in M. tb Lipidome MS-LAMP. Thus far, there are no haloge-nated M. tb lipids. Therefore, it can be understood from this tablethat for GLs, PKs, PRs, SPs and STs, there could be less ambiguitiesin differentiating the peaks due to most abundant elements ofthe molecular ion from the signals contributed by their isotopes.




Figure 5. Population distribution of different types of lipids across eight categories in M. tb Lipidome MS-LAMP: (a) M. tb lipidome, (b) all phospholipids,(c) sulfolipids and (d) pie chart representation of population distribution across 15 main classes: 1. fatty acids and conjugates (FA01); 2. fatty esters (FA07);3. monoacylglycerols (MG); 4. diacylglycerols (DG); 5. triacylglycerols (TG); 6. glycerophosphoethanolamines (PE); 7. glycerophosphoglycerols (GPG); 8.glycerophosphoglycerophospho-glycerols (CL); 9. glycerophosphoinositolglycans (GPIG); 10. glycerophosphoinositols (PI); 11. linear polyketides (PK01); 12.quinones and hydroquinones (PR02); 13. polyprenols (PR03); 14. acyltrehaloses (SL03); and 15. sulfolipids (SL03y) (Source:Mtb LipidDB, Mycobacteria ResearchLaboratories, Colorado State University: www.mrl.colostate.edu; as of September 2012).

Table 2. Number of lipids of known and unknown molecular mass inGeneral Lipidome MS-LAMP distributed across eight categories#

Category Total no.of lipids

No. of lipidsof unknown

molecular mass

No. of lipidsof known

molecular mass

Fatty acyls (FA) 5791 3 5788

Glycerolipids (GL) 7538 14 7524

Glycerophospholipids

(GP)

8005 22 7983

Polyketides (PK) 6744 2 6742

Prenol lipids (PR) 1200 20 1180

Saccharolipids (SL) 1293 0 1293

Sphingolipids (SP) 4320 435 3885

Sterol lipids (ST) 2681 20 2661

Total 37 572 516 37 056

# Source: LIPID MAPS consortium (www.lipidmaps.org) (updatedup to June 2012).

Table 3. Total number of lipids in General Lipidome MS-LAMP andtheir molecular mass range#

Molecular mass range (in Da) 46.00548–4722.081313

Total entries 37572

Unique entries (based on molecular

mass)

6844

# Source: LIPID MAPS consortium (www.lipidmaps.org) (updatedup to June 2012).



471

However, for 518 FAs (whose masses are greater than 1300Da),542 GPs and 189 SLs (whose masses are higher than ~1550Da),identifying signal(s) due to molecular ion peaks and their respec-tive isotope peaks, whose masses differ by 1 mass unit, can beconvoluted. Thus, it may be imperative that different querytypes and functionalities of MS-LAMP (especially QT1 and QT2)need to be utilized in conjunction, so as to arrive at appropriate



Table 4. Lipids having more than 90 carbon atoms distributed across eight categories in M. tb Lipidome MS-LAMP*

Category No. of lipids having morethan 90 carbon (C) atoms

Range of molecular mass of lipidshaving more than 90 C atoms

Molecular formula of lipidhaving maximum C atoms

Fatty acyls (FA) 581 1305.361945 – 3064.965305 C200 H390 O17 (200)

Glycerolipids (GL) 0 — C89 H174 O6 (89)

Glycerophospholipids (GP) 542 1585.191327 – 2411.458611 C120 H219 O45 P1 (120)

Polyketides (PK) 0 — C47 H74 N5 O10 Fe1 (47)

Prenol lipids (PR) 0 — C56 H82 O6 S1 (56)

Saccharolipids (SL) 189 1589.221355 – 2796.510135 C174 H338 O21 S1 (174)

Sphingolipids (SP) 0 — —

Sterol lipids (ST) 0 — —

* Source: Mycobacteria Research Laboratories, Colorado State University (www.mrl.colostate.edu) (as of September 2012).

Figure 6. Population distribution of different types of lipids across eight categories in General Lipidome MS-LAMP: (a) General Lipidome; (b) allphospholipids; (c) all phospholipids (barring GP): 527 lipids; (d) sulfur-containing lipids: 457 lipids; (e) halogenated lipids: 345 lipids (Source: LIPID MAPSconsortium: www.lipidmaps.org; last updated June 2012).


472

assignments of mass spectra towards identifying lipids unambig-uously and more accurately.In the case of General Lipidome MS-LAMP, differentiating

isotope peaks from the peaks of molecular ions consisting ofmost abundant elements can be complicated for 1253 SLs,1478 SPs and for a few GLs, GPs and PRs (Table 5). For SLs andSPs in this case, special attention may have to be given onlybeyond the mass of 1800Da to ensure proper assignmentsand identifications.


Discussion

Lipids have always been found to be interesting and importantin the contexts of biology, medicine and healthcare. Indeed,in disorders such as obesity, hypertension, diabetes andeven in the case of Alzheimer’s disease, a neurological disorder,implication of lipids have been realized.[7,9,11] Furthermore,characterization of lipidome or membrane lipidome of parasitescausing diseases such as malaria and leishmaniasis can be of



Table 5. Lipids having more than 90 carbon atoms distributed across eight categories in General Lipidome MS-LAMP#

Category No. of lipids having more than90 carbon (C) atoms

Range of molecular mass of lipidshaving more than 90 C atoms

Molecular formula of lipidhaving maximum C atoms

Fatty acyls (FA) 0 — C76 H142 O15 (76)

Glycerolipids (GL) 5 1463.368215–1933.417746 C104 H205 O26 P S (104)

Glycerophospholipids (GP) 4 1591.383817–1661.462067 C99 H202 O13 P2 (99)

Polyketides (PK) 1 1730.39599–1730.39599 C90 H74 O36 (90)

Prenol lipids (PR) 6 1245.153015–2109.42245 C122 H202 N2 O22 P2 (122)

Saccharolipids (SL) 1253 1809.25578–2282.04704 C146 H272 O16 (146)

Sphingolipids (SP) 1478 1860.085987–4722.081313 C198 H336 N12 O115 (198)

Sterol lipids (ST) 0 — C59 H100 O26 (59)

# Source: LIPID MAPS consortium (www.lipidmaps.org) (updated up to June 2012).

There are no lipids with more than 90 C atoms and chlorine, and 90 C atoms and bromine.


immense worth to comprehend the issues related with multi-drugresistant strains, and accordingly, strategies can be devised toovercome such resistances.[27,28] In this regard, characterization ofM. tb is also very important, mainly due to the increase in theemergence of new multi- and extensively drug resistant strains,particularly towards major drugs.[29] The physicochemical propertiesof the cell wall have a significant role in contributing to the virulenceof M. tb.[30,31] A very interesting aspect that strongly encouragesinvestigation of lipidome of M. tb is that nearly 60% of the cellulardry weight is lipids that could enhance their cell envelope’simpermeability.[23] Therefore, characterization of M. tb lipidome,especially cell surface lipids, would provide insights to understandabout host–pathogen interactions, thereby determine those lipidbiosynthetic pathways that might be responsible for pathogenesisand drug resistance. Accordingly, suitable drugs can be designed.

In vitro studies have shown that M. tb is more susceptible toacidic pH as compared with other mycobacterial species.[32,33]

Vulnerability ofM. tb (H37Ra in liquid medium) towards weak acids

Figure 7. ESI mass spectra of hexane-only extract (cell wall lipids) of M. tb g


at acidic pH has also been demonstrated.[33] However, in vivo, M. tbis capable to thrive and grow within acidic conditions presentedby the phagocytic vacuoles of host macrophages.[34,35] Globaltranscriptomic profiling of M. tb infecting human macrophage likeTHP-1 cells revealed main transcriptional alterations related tometabolism of lipids and remodelling of cell envelope.[36] Hence,examining variations of lipidome ofM. tb fromdifferent pH environ-ments can be of worth, whereby it is possible to determine themechanisms and factors that enable this pathogen to adapt forsurvival in acidic environment. Indeed, alterations of cell walllipids extracted from different strains, including that of someknocked-out strains of M. tb, grown in media of different pH, hasbeen shown.[35]

Thus, to evaluate and demonstrate the utility of MS-LAMPdeveloped herein, it was decided to acquire ESI mass spectrometricdata of lipids extracted from a laboratory strain of M. tb grown intwo different acidic environments, pH7.0 and 5.5 (see Materialsand Methods section).

rown in media maintained at (a) pH 7.0 and (b) pH 5.5.


473



474

Comparison of lipidome profiles of M. tb grown under pH7.0and 5.5

A laboratory strain of M. tb was grown under two different pHconditions, 7.0 and pH 5.5, and from these, two different lipidextracts were prepared: (i) total lipid extraction by chloroform(CHCl3)/methanol (CH3OH) (2 : 1; v/v) and (ii) extraction of cell walllipids by hexane only. These extracts were then subjected todirect infusion ESI–MS (see Materials and Methods section) toinvestigate the changes occurring to lipidome profiles on varyingthe pH. Figure 7 compares the ESI mass spectra of hexane-onlyextracts of M. tb grown in pH 7.0 and pH 5.5 media, whereasthe mass spectral data of CHCl3–CH3OH extracts are shown inFig. S7 (Supporting information). M. tb Lipidome MS-LAMP was

Table 6. Lipids identified from ESI mass spectrometric data† of two differ

Lipid Hexane

Category pH 7.0

1 Fatty acyls (FA) 28

2 Glycerolipids (GL) 0

3 Glycerophospholipids (GP) 4

4 Polyketides (PK) 0

5 Prenol lipids (PR) 0

6 Saccharolipids (SL) 5

The aforementioned results are the output of M. tb lipidome MS-LAMP on sThe other parameters queried in are as follows: ion type, [M�H]�; windoversion of M. tb Lipidome MS-LAMP, there are no lipids in the categoriescategories are not included in this table.

Table 7. Potential lipids identified from hexane-only extract of M. tb grow

Queriedm/z values†

Queried ion type

[M�H]�

Molecular formula Lipid Molecu

1355.99 — — C73 H140

1380.04 C75 H146 O17 P2 CL (GP)

1455.99* C75 H141 O24 P1,

C81 H150 O17 P2

Ac1PIM2 (GP), CL (GP) C78 H150

1480.03 — — C80 H150

1555.98* — —

1580.03 C83 H153 O25 P1 Ac2PIM2 (GP)

1655.98* C84 H153 O29 P1 Ac1PIM3 (GP)

1680.03 — — C83 H153

C88 H157

1755.97* C86 H149 O34 P1 Ac1PIM4 (GP) C83 H149

1780.02 C87 H161 O34 P1 Ac1PIM4 (GP)

1855.97* C87 H157 O39 P1 Ac1PIM5 (GP)

1880.03 C89 H157 O39 P1 Ac1PIM5 (GP)

1980.02 — —

† These m/z values correspond to some of the intense signals in range m/z1383.01, 1483.01, 1583.01, 1683.00, 1783.00, 1883.00, 1982.99 and 2082.99LAMP database, on using the aforementioned queried ion types and wind

* These m/z values are also observed in ESI mass spectrum shown in Fig.} Same identifications for different m/z values queried using different quer

CL, cardiolipin; AcPIM, acylated glycerophosphoinositol mannoside; Ac2SG


useful to investigate the variations in the lipidome profile. Fromthese mass spectra (Figs 7 and S7), m/z values were manuallychosen and filtered on the basis of a defined threshold ofsignal-to-noise ratio. These values were then uploaded inthe form of a .txt file in query type 2 of M. tb Lipidome MS-LAMPand lipids corresponding to this data were identified (Fig. S2,Supporting information, shows the output corresponding to datarecorded of hexane-only extract obtained from pH7.0 medium).Table 6 compares different categories of lipids determinedfrom ESI mass spectra (Figs 7 and S7) of lipid extracts obtainedfrom M. tb grown in pH 7.0 and 5.5.

The mass spectrum acquired from hexane extract of M. tbgrown in pH 5.5 contains several intense peaks in the range

ent solvent extracts obtained from M. tb grown in pH 7.0 and 5.5

only (cell wall lipids) CHCl3–CH3OH (total lipids)

pH 5.5 pH 7.0 pH 5.5

25 0 7

3 0 1

66 55 35

0 0 0

0 0 0

1 6 1

ubmitting ESI–MS data (Figs 7† and S7†) using query type 2 (Fig. S2†).w m/z, �0.05; and lipids, all eight categories (Fig. S2†). In the currentof sterols and sphingolipids (Fig. 5 and Table S1), and therefore, these

n in pH5.5 (using M. tb Lipidome MS-LAMP)

(window range m/z, �0.05)

[M+Cl]� [M+HOAc�H]-

lar formula Lipid Molecular formula Lipid

O17 S1} Ac2SGL (SL)} — —

— — C73 H140 O17 S1} Ac2SGL (SL)}

O17 P2} CL (GP)} — —

O17 P2 CL (GP) C78 H150 O17 P2} CL (GP)}

— — C78 H145 O24 P1 Ac1PIM2 (GP)

— — — —

— — C80 H141 O29 P1 Ac1PIM3 (GP)

O29 P1,

O25 P1

Ac1PIM3 (GP),

Ac2PIM2 (GP)

— —

O34 P1 Ac1PIM4 (GP) C81 H149 O34 P1 Ac1PIM4 (GP)

— — C83 H149 O34 P1 Ac1PIM4 (GP)

— — — —

— — — —

— — C92 H161 O39 P1 Ac1PIM5 (GP)

1300–2000 in ESI mass spectrum shown in Fig. 7(b). The m/z values,(Fig. 7(b) and Table S3) do not match to any lipid inM. tb lipidome MS-ow range m/z from �0.005 up to �0.25.

S7b (Supporting information; CHCl3–CH3OH extract from pH5.5).

ied ion types.

L, acylated sulfolipid.


Figure 8. ESI-MS/MS spectra of two mycolic acids recorded in negative ion polarity using [M�H]� precursor ion (see Materials and Methods section formore details): (a) precursor ion m/z 1136.18; (b) precursor ion m/z 1164.23; (c) summary of interpretations of the MS/MS spectra shown in (a) and (b).


m/z 1300–2000 (Fig.7(b)), which are absent in the spectrumrecorded from the hexane extract of M. tb grown in pH 7.0(Fig. 7(a)). M. tb lipidome MS-LAMP was used to identifypotential lipids that could contribute to some of the intensesignals in this m/z range, and it was noted that many ofthese are GPs (Table 7). Further, a few peaks of good intensityat m/z 1383.01, 1483.01, 1583.01, 1683.00, 1783.00, 1883.00,1982.99 and 2082.99 (Fig. 7(b)) did not correspond to any lipidin M. tb lipidome, as found by using three different iontypes ([M�H]�, [M +Cl]� and [M+HOAc�H]�) and varyingwindow range m/z from �0.005 to �0.25; with an exceptionof one case: m/z 2082.99 can be [M+HOAc�H]� of a glycosy-lated phthiocerol dimycocerosate (C125 H234 O18), which wasobtained on selecting window range m/z �0.25. On usingGeneral Lipidome MS-LAMP, when the same three ion types(as mentioned earlier) were chosen and window range m/zvaried up to �0.5, these eight m/z values predominantlycorresponded to SPs (Table S3, Supporting information), whichare currently absent in M. tb Lipidome (Fig. 5 and Table S1).Perhaps, in M. tb, these are new lipids and/or SPs, whichneed to be elucidated. Probably, the acidic environment(pH 5.5) is responsible for triggering the syntheses of GPs,new lipids and/or SPs (in mass range m/z 1300–2000), whichmay not be produced preponderantly by mycobacterium inpH 7 condition.

475

Tandem mass spectrometric identification of somemycolic acids

The lipid anions detected at m/z 1136.20 and 1164.24 (Fig. 7) werechosen for MS/MS experiments through CID (see Materials andMethods section). Prior to MS/MS analyses, these two lipids werefound to belong to the FA category, as searched using M. tblipidome MS-LAMP. Specifically, these are alpha-mycolic acids,


C78 H152 O3 (monoisotopic molecular mass 1137.174145Da)and C80 H156 O3 (monoisotopic molecular mass 1165.205445Da).Hence, the peaks at m/z 1136.20 and 1164.24 correspond to singlydeprotonated form [M�H]�. The ESI–MS/MS spectra of these twoprecursor ions recorded from hexane and CHCl3–CH3OH extractswere found to be identical to MS/MS spectra of the same two lipidprecursor anions that were acquired from a commercial standard(Sigma, M4537) (Fig. 8). The major product ion at m/z 395.40 canbe interpreted as to a-branch of mycolic acid: C26 H51 O2. Themassdifference between m/z values of precursor ion and a-branchmatches to the mass of meromycolate moiety (Fig. 8). Recently,several mycolic acids and their isomers possessing different chainlengths of a- and meromycolate branches were identified from apathogenic bacteria Rhodococcus equi using negative ion ESI CIDMS/MS.[37] From Fig. 8, an isomer of mycolic acid, C78 H152 O3(monoisotopic molecular mass 1137.174145Da) having a-branchof a different chain length, C24 H47 O2 can be realized, as evidentfrom the product ion m/z 367.37.

Conclusion

It is well known that computational exercises and analyses oflipidome cannot be carried out in a manner similar to that ofgenome and proteome. Atoms, molecules and molecularstructures are perhaps the only variables used for defining lipidsand to perform computational analyses of lipidome. The diversityof molecular structures of lipids is so enormous and complicatedthat classifying-cum-naming them in a systematic manner isessential. Carrying out this task meticulously can enable thesearch process of tracing one or more (specific) lipid(s) amidsthuge population somewhat easier. Nevertheless, many a times,it is not very straightforward to search, even after building well-planned and ordered databases. And therefore, different toolsof searching are developed with a view to extract appropriate



476

information from a vast set. In doing so, the tools can follow avariety of ways to survey the databases and pull out the informationthat a user desires for. One of the ways to search and obtain infor-mation on lipids is by using mass spectrometric data. Alternatively,molecular formula also can be utilized for searching lipids.The LIPID MAPS consortium (www.lipidmaps.org) has a compre-

hensive collection of lipids that has been classified in a very method-ical fashion. The eight-category classification has aided in reducingthe complexity and in improving the understanding of molecularstructures of lipids. AWeb tool and a freely downloadable standaloneversion (LIPID MS Predict) are available, which can be of use tosearch lipids utilizing mass spectrometric data. However, severalfeatures are unique to MS-LAMP making it more flexible.Some advantages of MS-LAMP are as follows: (i) standalone

(non Web-based); (ii) output representation providing a quick lookat all categories of lipids corresponding to a set of experimentaldata submitted; (iii) any database can be integrated into theprogram, however the database will need to be modified suitableto the design of the program; (iv) wide varieties of choices to directthe search in a more specific manner, which includes newlyintroduced lipid groups and diverse ion types (of positive andnegative ions); and (v) experimental data can be submitted as a textfile containing list ofm/z values. The output representation provid-ing a glimpse of different lipid categories is a highlighting aspectof MS-LAMP. Such a representation can aid in readily eliminatingcertain hits and focus on the more prospective candidates. Becauseany database can be integrated, the program can be customizedaccording to the requirements or objectives of the research project.Furthermore, a variety of data file formats obtained frominstruments of different manufacturers can be readily convertedto text file containing a list of m/z and their respective intensityvalues. Therefore, MS-LAMP can be used to work with the dataacquired from a wide range of instruments. Altogether, thesemerits facilitate in providing a broad overview of results, wherebyglimpse of outputs from multiple databases can be comparedusing a single platform. This implies, for a given experimentaldata, the outputs yielding from databases of different biologicalspecies can be examined and verified, through which it would bepossible to know commonalities and variations across differentspecies/organisms in terms of their respective lipidomes. Further-more, MS-LAMP can be used in Linux as well, in which case PyMOLcan be used to view molecular structures that are obtained in theoutput from General Lipidome MS-LAMP.This software was also conceived to facilitate clearer under-

standing of the mass spectral data of lipids. And therefore, itindeed possesses a role as a very useful learning tool for novices,who begin to utilize MS for studying lipids from diverse biologicalsources. All these features clearly indicate that MS-LAMP isdistinct from the already existing software tools to interpret massspectral data of lipids.MS-LAMP can be particularly useful for top-down lipidomic

studies involving high resolution mass spectrometric data. Inparticular, MS-LAMP can be valuable for comparative targetedanalyses, where qualitative lipid profiling, i.e. lipid massfingerprinting, is sufficient to gain preliminary insights precedingin-depth investigations. In other words, MS-LAMP shortlistspotential lipids of relevance for further studies, and therefore,quantitative characterizations can be focused on only thoselipids. Accordingly, appropriate standards (chemically synthe-sized) can be selected. Thus, the standards that might beirrelevant under some given circumstances can be obviatedthrough MS-LAMP. Further, to address a variety of lipidomic


investigations, particularly related to tuberculosis, this standaloneversion will be of immense utility.

Acknowledgements

V. S. is thankful to Dr Rajesh S. Gokhale, Director, CSIR-IGIB, andMs Priyanka Verma, PhD Research Scholar, National Institute ofImmunology, New Delhi, India, for kindly providing lipid extractsderived of a laboratory strain of M. tb. V. S. wishes to thankDr Beena Pillai, Dr Shantanu Sengupta and Dr Sridhar Sivasubbu,Scientists, CSIR-IGIB, for providing various kinds of support towardsproteomics lab, where mass spectrometric experiments were con-ducted. Mr M. V. Raghu Nandanan, IT Division, CSIR-IGIB, is thankedfor making arrangements towards Web hosting. The projectsMLP6201, FAC002, NWP036 and NWP004 of CSIR, Government ofIndia, are gratefully acknowledged for their generous funding.

A part of this study began as a 6-month industrial training ofG. S. carried out at CSIR-IGIB under the guidance of V. S. Thepreliminary results of this work comprising initial efforts andanalyses made towards making this software were presented asan oral presentation in the National Conference on Biopharmaceu-ticals & Healthcare (November 4th & 5th, 2011) organized byDepartment of Biotechnology, Sir M. Visvesvaraya Institute ofTechnology, Bengaluru – 562157, India.

Supporting information

Supporting information may be found in the online version ofthis article.

References[1] H. A. Brown, R. C. Murphy. Working towards an exegesis for lipids in

biology. Nat. Chem. Biol. 2009, 5(9), 602.[2] S. J. Blanksby, T. W. Mitchell. Advances in mass spectrometry for

lipidomics. Annu.Rev.Anal.Chem.(Palo.Alto.Calif.). 2010, 3, 433.[3] K. M. Bou, W. Hou, H. Zhou, F. Elisma, L. A. Swayne, A. P. Blanchard,

Z. Yao, S. A. Bennett, D. Figeys. Lipidomics era: accomplishmentsand challenges. Mass Spectrom. Rev. 2010, 29(6), 877.

[4] W. J. Griffiths. Tandem mass spectrometry in the study of fatty acids,bile acids, and steroids. Mass Spectrom. Rev. 2003, 22(2), 81.

[5] F. Gonzalez-Illan, G. Ojeda-Torres, L. M. Diaz-Vazquez, O. Rosario.Detection of fatty acid ethyl esters in skin surface lipids as biomarkersof ethanol consumption in alcoholics, social drinkers, light drinkers,and teetotalers using a methodology based on microwave-assistedextraction followed by solid-phase microextraction and gaschromatography-mass spectrometry. J. Anal. Toxicol. 2011, 35(4), 232.

[6] N. J. Jensen, K. B. Tomer, M. L. Gross. FABMS/MS for phosphatidylinositol,-glycerol, -ethanolamine and other complex phospholipids. Lipids1987, 22(7), 480.

[7] X. Han, M. Holtzman, D. W. McKeel, Jr., J. Kelley, J. C. Morris.Substantial sulfatide deficiency and ceramide elevation in very earlyAlzheimer’s disease: potential role in disease pathogenesis. J. Neuro-chem. 2002, 82(4), 809.

[8] X. Han, H. Cheng. Characterization and direct quantitation of cere-broside molecular species from lipid extracts by shotgun lipidomics.J. Lipid Res. 2005, 46(1), 163.

[9] J. Graessler, D. Schwudke, P. E. Schwarz, R. Herzog, A. Shevchenko,S. R. Bornstein. Top-down lipidomics reveals ether lipid deficiency inblood plasma of hypertensive patients. PLoS One 2009, 4(7), e6261.

[10] O. Quehenberger, A. M. Armando, A. H. Brown, S. B. Milne, D. S. Myers,A. H. Merrill, S. Bandyopadhyay, K. N. Jones, S. Kelly, R. L. Shaner, C. M.Sullards, E. Wang, R. C. Murphy, R. M. Barkley, T. J. Leiker, C. R. Raetz,Z. Guan,G.M. Laird, D. A. Six, D.W. Russell, J. G.McDonald, S. Subramaniam,E. Fahy, E. A. Dennis. Lipidomics reveals a remarkable diversity oflipids in human plasma. J. Lipid Res. 2010, 51(11), 3299.

[11] X. Han, S. Rozen, S. H. Boyle, C. Hellegers, H. Cheng, J. R. Burke,K. A. Welsh-Bohmer, P. M. Doraiswamy, R. Kaddurah-Daouk.Metabolomics in early Alzheimer’s disease: identification of altered




plasma sphingolipidome using shotgun lipidomics. PLoS One2011, 6(7), e21643.

[12] R. Herzog, D. Schwudke, K. Schuhmann, J. L. Sampaio, S. R. Bornstein,M. Schroeder, A. Shevchenko. A novel informatics concept for high-throughput shotgun lipidomics based on themolecular fragmentationquery language. Genome Biol. 2011, 12(1), R8.

[13] R. Herzog, K. Schuhmann, D. Schwudke, J. L. Sampaio, S. R. Bornstein,M. Schroeder, A. Shevchenko. LipidXplorer: a software for consensualcross-platform lipidomics. PLoS One 2012, 7(1), e29851.

[14] K. Yang, H. Cheng, R. W. Gross, X. Han. Automated lipid identificationand quantification by multidimensional mass spectrometry-basedshotgun lipidomics. Anal. Chem. 2009, 81(11), 4356.

[15] H. Song, F. F. Hsu, J. Ladenson, J. Turk. Algorithm for processing rawmass spectrometric data to identify and quantitate complex lipid mo-lecular species in mixtures by data-dependent scanning and fragmention database searching. J. Am. Soc. Mass Spectrom. 2007, 18(10), 1848.

[16] D. Schwudke, J. Oegema, L. Burton, E. Entchev, J. T. Hannich, C. S. Ejsing,T. Kurzchalia, A. Shevchenko. Lipid profiling by multiple precursor andneutral loss scanning driven by the data-dependent acquisition. Anal.Chem. 2006, 78(2), 585.

[17] C. S. Ejsing, E. Duchoslav, J. Sampaio, K. Simons, R. Bonner, C. Thiele,K. Ekroos, A. Shevchenko. Automated identification and quantificationof glycerophospholipid molecular species by multiple precursor ionscanning. Anal. Chem. 2006, 78(17), 6202.

[18] P. Haimi, A. Uphoff, M. Hermansson, P. Somerharju. Software toolsfor analysis of mass spectrometric lipidome data. Anal. Chem.2006, 78(24), 8324.

[19] M. D. Leavell, J. A. Leary. Fatty acid analysis tool (FAAT): an FT-ICR MSlipid analysis algorithm. Anal. Chem. 2006, 78(15), 5497.

[20] G. Hubner, C. Crone, B. Lindner. lipID–a software tool for automatedassignment of lipids inmass spectra. J. Mass Spectrom. 2009, 44(12), 1676.

[21] T. Pluskal, S. Castillo, A. Villar-Briones,M.Oresic. MZmine 2:modular frame-work for processing, visualizing, and analyzing mass spectrometry-based molecular profile data. BMC Bioinformatics 2010, 11, 395.

[22] E. M. Hein, B. Bodeker, J. Nolte, H. Hayen. Software tool for miningliquid chromatography/multi-stage mass spectrometry data forcomprehensive glycerophospholipid profiling. Rapid Commun. MassSpectrom. 2010, 24(14), 2083.

[23] M. J. Sartain, D. L. Dick, C. D. Rithner, D. C. Crick, J. T. Belisle. Lipidomicanalyses of Mycobacterium tuberculosis based on accurate massmeasurements and the novel “Mtb LipidDB”. J. Lipid Res. 2011, 52(5), 861.

[24] E. Fahy, S. Subramaniam, H. A. Brown, C. K. Glass, A. H. Merrill, Jr.,R. C. Murphy, C. R. Raetz, D. W. Russell, Y. Seyama, W. Shaw, T. Shimizu,


F. Spener, M. G. van, M. S. VanNieuwenhze, S. H. White, J. L.Witztum, E. A. Dennis. A comprehensive classification system for lipids.J. Lipid Res. 2005, 46(5), 839.

[25] E. Fahy, S. Subramaniam, R. C. Murphy, M. Nishijima, C. R. Raetz,T. Shimizu, F. Spener, M. G. van, M. J. Wakelam, E. A. Dennis. Updateof the LIPID MAPS comprehensive classification system for lipids.J. Lipid Res. 2009, 50 (Suppl), S9.

[26] M. Sud, E. Fahy, D. Cotter, A. Brown, E. A. Dennis, C. K. Glass,A. H. Merrill, Jr., R. C. Murphy, C. R. Raetz, D. W. Russell, S. Subramaniam.LMSD: LIPID MAPS structure database. Nucleic Acids Res. 2007,35(Database issue), D527–D532.

[27] M. C. Ben, S. T. Prigge, H. Vial. Targeting the lipid metabolic pathwaysfor the treatment of malaria. Drug Dev. Res. 2010, 71(1), 44.

[28] E. Marechal, M. Riou, D. Kerboeuf, F. Beugnet, P. Chaminade, P.M. Loiseau.Membrane lipidomics for the discovery of new antiparasitic drugtargets. Trends Parasitol. 2011, 27(11), 496.

[29] C.Y. Chiang, R. Centis, G. B. Migliori. Drug-resistant tuberculosis: past,present, future. Respirology 2010, 15(3), 413.

[30] C. E. Barry, III. Interpreting cell wall ‘virulence factors’ of Mycobacteriumtuberculosis. Trends Microbiol. 2001, 9(5), 237.

[31] M. Jain, C. J. Petzold, M. W. Schelle, M. D. Leavell, J. D. Mougous,C. R. Bertozzi, J. A. Leary, J. S. Cox. Lipidomics reveals control ofMycobacterium tuberculosis virulence lipids via metabolic coupling.Proc. Natl. Acad. Sci. U. S. A. 2007, 104(12), 5133.

[32] D. L. Piddington, A. Kashkouli, N. A. Buchmeier. Growth of Mycobacte-rium tuberculosis in a defined medium is very restricted by acid pHand Mg(2+) levels. Infect. Immun. 2000, 68(8), 4518.

[33] Y. Zhang, H. Zhang, Z. Sun. Susceptibility ofMycobacterium tuberculosisto weak acids. J. Antimicrob. Chemother. 2003, 52(1), 56.

[34] K. Rohde, R. M. Yates, G. E. Purdy, D. G. Russell. Mycobacteriumtuberculosis and the environment within the phagosome. Immunol.Rev. 2007, 219, 37.

[35] R. B. Abramovitch, K. H. Rohde, F. F. Hsu, D. G. Russell. aprABC: aMycobacterium tuberculosis complex-specific locus that modulatespH-driven adaptation to the macrophage phagosome. Mol. Microbiol.2011, 80(3), 678.

[36] P. Fontan, V. Aris, S. Ghanny, P. Soteropoulos, I. Smith. Globaltranscriptional profile of Mycobacterium tuberculosis during THP-1human macrophage infection. Infect. Immun. 2008, 76(2), 717.

[37] F. F. Hsu, K. Soehl, J. Turk, A. Haas. Characterization of mycolicacids from the pathogen Rhodococcus equi by tandem massspectrometry with electrospray ionization. Anal. Biochem. 2011,409(1), 112.


477

Mass spectrometry based lipid(ome) analyzer and molecular platform: a new software to interpret and...

Documents

Transcript of Mass spectrometry based lipid(ome) analyzer and molecular platform: a new software to interpret and...