Abstract

A major challenge of lipidomics is to determine and quantify the precise content of complex lipidomes to the exact lipid molecular species. Often, multiple methods are needed to achieve sufficient lipidomic coverage to make these determinations. Multiplexed targeted assays offer a practical alternative to enable quantitative lipidomics amenable to quality control standards within a scalable platform. Herein, we developed a multiplexed normal phase liquid chromatography-hydrophilic interaction chromatography multiple reaction monitoring method that quantifies lipid molecular species across over 20 lipid classes spanning wide polarities in a single 20-min run. Analytical challenges such as in-source fragmentation, isomer separations, and concentration dynamics were addressed to ensure confidence in selectivity, quantification, and reproducibility. Utilizing multiple MS/MS product ions per lipid species not only improved the confidence of lipid identification but also enabled the determination of relative abundances of positional isomers in samples. Lipid class-based calibration curves were applied to interpolate lipid concentrations and guide sample dilution. Analytical validation was performed following FDA Bioanalytical Method Validation Guidance for Industry. We report repeatable and robust quantitation of 900 lipid species measured in NIST-SRM-1950 plasma, with over 700 lipids achieving inter-assay variability below 25%. To demonstrate proof of concept for biomarker discovery, we analyzed plasma from mice treated with a glucosylceramide synthase inhibitor, benzoxazole 1. We observed expected reductions in glucosylceramide levels in treated animals but, more notably, identified novel lipid biomarker candidates from the plasma lipidome. These data highlight the utility of this qualified lipidomic platform for enabling biological discovery.

Graphical abstractSupplementary Key wordsAbbreviations: BA (bioanalytical), BSA (bovine serum albumin), Cer (ceramide), DAG (diacylglycerol), DB (double bond), DCM (dichloromethane), FA (fatty acyl), FDA (food and drug administration), GalCer (galactosyl-ceramide), GCS (glucosylceramide synthase), GlcCer (glucosyl-ceramide), HexCer (hexosylceramide), HILIC (hydrophilic interaction chromatography), IPA (2-propanol), IS (internal standard), ISF (in-source fragmentation), LOQ (limit of quantitation), NPLC (normal phase chromatography), PC (phosphatidylcholine), PE (phosphatidylethanolamine), PG (phosphatidylglycerol), QqQ (triple quadrupole), RT (retention time), SIL (stable isotope labeled), SM (sphingomyelin), UHPSFC (ultrahigh-performance supercritical fluid chromatography), ULOQ (upper limits of quantitation)Lipids are vital biomolecules serving fundamental roles in cellular signaling, cell membrane architecture, energy storage, and metabolism. Seemingly minor structural differences among individual lipid species, such as the number, position, and geometry of double bonds (DBs) in acyl chains, are pivotal determinants of their functions (1Róg T. Orłowski A. Llorente A. Skotland T. Sylvänne T. Kauhanen D. et al.Interdigitation of long-chain sphingomyelin induces coupling of membrane leaflets in a cholesterol dependent manner.). Alterations in lipid metabolic network may trigger a cascade of deleterious cellular events. The increasing evidence of the biological relevance of lipids and their roles in diseases such as neurodegenerative (2Fanning S. Haque A. Imberdis T. Baru V. Barrasa M.I. Nuber S. et al.Lipidomic analysis of α-synuclein neurotoxicity identifies stearoyl CoA desaturase as a target for parkinson treatment.), cancer (3Veglia F. Tyurin V.A. Blasi M. De Leo A. Kossenkov A.V. Donthireddy L. et al.Fatty acid transport protein 2 reprograms neutrophils in cancer.), cardiovascular (4Laaksonen R. Ekroos K. Sysi-Aho M. Hilvo M. Vihervaara T. Kauhanen D. et al.Plasma ceramides predict cardiovascular death in patients with stable coronary artery disease and acute coronary syndromes beyond LDL-cholesterol.), and infectious diseases (5Song J.W. Lam S.M. Fan X. Cao W.J. Wang S.Y. Tian H. et al.Omics-driven systems interrogation of metabolic dysregulation in COVID-19 pathogenesis.) during the past decade, has expanded the development of lipidomics for use in biomedical research (6Holčapek M. Liebisch G. Ekroos K. ). Analytical methods that accurately quantify lipid molecular species in biological samples are pivotal to enable understanding of underlying biological mechanisms of disease pathology at a deeper level and offer a potential to accelerate the development of new therapeutics.Lipids have been classified into eight major categories in the LIPID MAPS repository (7Fahy E. Subramaniam S. Murphy R.C. Nishijima M. Raetz C.R.H. Shimizu T. et al.Update of the LIPID MAPS comprehensive classification system for lipids.) based on their building blocks. From analytical and physiochemical standpoints, neutral lipids are hydrophobic molecules lacking charged groups. In contrast, polar lipids have variety of polar headgroups. Both types share the complexities in fatty acyl (FA) chains including but not limited to chain lengths and positions of DBs. In addition, a large variation in stereochemistry also exists, such as epimers (e.g., glucosyl- and galactosyl-ceramide (GlcCer and GalCer)) and regioisomers (e.g., sn-1 and sn-2 acyl position of phospholipids). This structural heterogeneity generates a large number of distinct species, which contributes to an extremely complex lipidomic chemical space (8Porta Siegel T. Ekroos K. Ellis S.R. Reshaping lipid biochemistry by pushing barriers in structural lipidomics.). Eukaryotic cells may contain thousands of distinct species with their lipid concentrations varying between organelles up to several million fold (). All these factors pose extreme challenges to the analytics required to achieve precise elucidation and quantification of distinct lipid species.Despite significant technological and methodological advances over the past decade, determining the precise content of a complex lipidome at the level of exact molecular species remains challenging. For example, lipid extraction methods that not only generate clean, enriched samples benefiting MS detection but also protect lipids from degradation and oxidation are needed (10Lipidomics, en route to accurate quantitation.). Similarly, there is a need for more confident data processing solutions especially for untargeted lipidomics data. The availability of lipid standards, especially stable isotope labeled (SIL) standards, determines the buildup of quantitative assays. In fact, standardization in lipidomics methods (11Liebisch G. Ahrends R. Arita M. Arita M. Bowden J.A. Ejsing C.S. et al.Lipidomics needs more standardization.) and lipid nomenclature (12Liebisch G. Fahy E. Aoki J. Dennis E.A. Durand T. Ejsing C.S. et al.Update on LIPID MAPS classification, nomenclature, and shorthand notation for MS-derived lipid structures.) has been initiated to address the discrepancies in quantitative information of existing methodologies and lipid annotation (13Liebisch G. Ekroos K. Hermansson M. Ejsing C.S. Reporting of lipidomics data should be standardized.). Typically, quantification is performed using one nonendogenous internal standard (IS) per lipid subclass (14Ejsing C.S. Duchoslav E. Sampaio J. Simons K. Bonner R. Thiele C. et al.Automated identification and quantification of glycerophospholipid molecular species by multiple precursor ion scanning.). A requirement and standard practice of industrial bioanalytical methods is to interpolate unknown concentrations against valid calibration curves. This is a necessity along with preset acceptance criteria in quality control samples according to the Food and Drug Administration (FDA) Bioanalytical (BA) Validation Guidance (15Food and Drug Administration
Bioanalytical Method Validation Guidance.) to ensure robust quantitation. In an idealized assay, a practical analytical workflow would align with ongoing lipidomics standardization (11Liebisch G. Ahrends R. Arita M. Arita M. Bowden J.A. Ejsing C.S. et al.Lipidomics needs more standardization.) to assure high quality data reported in a consistent manner.To date, quantitative lipidomics is achieved by either direct infusion or chromatography-based MS approaches. It is widely recognized that coionization of endogenous analytes and their ISs must be accounted for using MS-based quantification (16Ekroos K. V Chernushevich I. Simons K. Shevchenko A. Quantitative profiling of phospholipids by multiple precursor ion scanning on a hybrid quadrupole time-of-flight mass spectrometer.). Shotgun lipidomics satisfies this norm of quantification, and with its steady performance and high throughput, it becomes attractive in many ways (17Heiskanen L.A. Suoniemi M. Ta H.X. Tarasov K. Ekroos K. Long-term performance and stability of molecular shotgun lipidomic analysis of human plasma samples., 18Schuhmann K. Moon H.K. Thomas H. Ackerman J.M. Groessl M. Wagner N. et al.Quantitative fragmentation model for bottom-up shotgun lipidomics.). However, the simultaneous injection of lipid ions by direct infusion complicates the ionization process resulting in compromised detection of ions with either low abundances or weak ionization efficiencies. MS spectra interpretation is challenging and limited because of a high degree of overlapping signals. Chromatography separation can effectively reduce this complexity. Reversed-phase liquid chromatography provides superior separation (19Sandra K. Pereira A. dos S. Vanhoenacker G. David F. Sandra P. Comprehensive blood plasma lipidomics by liquid chromatography/quadrupole time-of-flight mass spectrometry.) based on the apolar properties residing in FA chains. However, it is not preferred for quantification, mostly due to the lack of SIL lipid ISs with various FA chains. Hydrophilic interaction chromatography (HILIC) and normal phase chromatography (NPLC) separate lipids primarily based on lipid classes. This enables a quantification strategy using a few representative lipid standards and SIL ISs. Still, HILIC and NPLC have separation limitations as HILIC is relatively less capable of resolving nonpolar lipids while NPLC is suboptimal in resolving polar lipids within a relatively short eluting period (e.g., in less than 15–20 min).Here, we seek an easily implementable workflow in line with FDA BA guidance and ongoing lipidomics standardization to generate quantitative lipidomics data to enable acceptable turnaround of robust lipidomic data sets. The method couples multiplexed NPLC-HILIC separation and fast scanning triple quadrupole (QqQ) MS employing Multiple Reaction Monitoring (MRM) detection within a 20-min run time per sample. The method is capable of providing deep structural and accurate quantitative information of lipidome content with resolution down to positional isomers of molecular species (8Porta Siegel T. Ekroos K. Ellis S.R. Reshaping lipid biochemistry by pushing barriers in structural lipidomics.). Applying this method, we report lipidome-wide alterations in a highly specific and well-studied pharmacological model by analyzing plasma of mice treated with a potent glucosylceramide synthase (GCS) inhibitor Benzoxazole 1 (BZ1). These data suggest that alterations of single points within lipid metabolic pathways can have broad effects across the lipidome that remain to be explored.Materials and methodsChemicals, reagents, and materialsHPLC-grade water, acetonitrile, methanol, chloroform, hexane, acetone, and dichloromethane (DCM) were obtained from Thermo Fisher Scientific (Fair Lawn, NJ). Formic acid, acetic acid, ammonium acetate, 2,6-Di-tert-butyl-4-methyphenol, 2-propanol (IPA), and bovine serum albumin (BSA) were purchased from Sigma-Aldrich (St. Louis, MO). SRM 1950 human plasma was obtained from National Institute of Standards and Technology (Gaithersburg, Maryland). Most lipid standards (supplemental Table S1) were products of Avanti Polar Lipids (Alabaster, AL) at the time of the experiment. 1x PBS buffer was prepared using PBS tablets from MiliporeSigma (Burlington, MA). BZ1 was synthesized by MSD chemists. Wheaton 4-ml and 20-ml amber glass vials with PTFF liner screw caps were used for lipid working solution preparations. Lipid extraction plate (2.0 ml glass conical insert in 96-well vial loader) was obtained from Chrom Tech (Apple Valley, MN). LC/MS injection plate (amber 1 ml low carry tapered glass inserts in 96-well plate with pre-slit cap mat) was from Analytical Sales & Services, Inc (Flanders, NJ). Various Rainin single and multi-channel pipettes and matching pipette tips were used for all manual liquid aliquoting. A Microlab Nimbus workstation equipped with a CORE 96 Probe Head (Hamilton, Reno, Nevada) was used for lipid extraction. A 96-well solvent evaporator device SPE Dry-96 was obtained from Jones Chromatography (Lakewood, CO) and connected to an in-house Nitrogen source. A Multi-Tube vortexer from VRM Scientific (Radnor, PA) was used for sample mixing.Inhibition of GCS in miceHeterozygous GBA1 D409V mice (C57BL/6N-Gba tm1.1 Mjff/J, stock number: 019106, Jackson Laboratory, Bar Harbor, ME) were rederived, bred, and maintained at Taconic. Animals were acclimated to housing for at least one week from delivery and kept on a normal 12 h/12 h light/dark cycle for a week or longer during which time access to food and water were provided ad libitum. Mice were divided into separate groups and fed chow formulated with a potent GCS inhibitor BZ1 or matching control chow for four days or four weeks. BZ1 content in chow (Rodent diet with 10 kcal% Fat and 220 mg of BZ1 API/kg; Research Diets, Inc., New Brunswick, NJ) was selected to target chronic exposure levels of unbound BZ1 approximating two times the reported GCS enzyme IC50 value of 18 nM (20Cosden M. Jinn S. Yao L. Gretzula C.A. Kandebo M. Toolan D. et al.A novel glucosylceramide synthase inhibitor attenuates alpha synuclein pathology and lysosomal dysfunction in preclinical models of synucleinopathy.). Weight of chow consumed was measured daily across individuals to estimate BZ1 daily doses and to verify similar feeding rates between mice-fed control versus BZ1-formulated diets. At the conclusion of treatment, animals were anesthetized with isoflurane (1–4% in oxygen at 2 l/minute) and blood was collected via cardiac puncture immediately prior to euthanasia via decapitation. Plasma was prepared and stored at −70°C until lipid extraction and analyses. All animal studies were performed under the approval of the Merck & Co., Inc., Kenilworth, NJ, Institutional Animal Care and Use Committee.Calibration standards and internal standards preparationA lipid standard mixture and a respective deuterated IS mixture were prepared. Ten calibration working solutions were prepared by serial diluting the standard mixture in DCM:IPA (1:1, v/v). The calibration standards were prepared by spiking the calibration working solutions into 2% BSA according to supplemental Tables S3 and S4 for method validation. The calibration curves were constructed by nominal concentrations of lipid calibrators (X-axis) and MS peak area ratios of the lipid standard over IS (Y-axis).Lipid extractionTo 20 μl mice plasma study samples, calibrators, and 2% BSA (single blanks), 10 μl IS mixture was spiked (supplemental Table S4). Ten microliters DCM:IPA (1:1, v/v) was spiked to several 20 μl 2% BSA solutions as the double blanks. Two hundred microliter ice-cold methanol containing 0.1% 2,6-Di-tert-butyl-4-methyphenol was added to all samples. The plate was mixed by vortexing gently for 2 min. Hamilton Microlab Nimbus was programmed to dispense 450 μl chloroform and 120 μl 20 mM acetic acid into each of the wells. The plate was mixed vigorously for 30 min. Phase separation was performed by centrifugation (Sorvall Legend centrifuge, Kendro Laboratory, Germany) at 2500 rpm for 10 min at 15°C. Three hundred sixty microliters of lower phase was transferred to a 1 ml glass LC/MS injection plate and dried under nitrogen gas at room temperature. To the remaining upper aqueous phase, 360 μl of chloroform was added and repeated as above. After centrifugation, the lower organic phase was pooled with the previous organic fraction. The final sample extracts were dried under nitrogen and reconstituted with 150 μl IPA:DCM (1:1, v/v) prior to analyses.LC-MRM analysisThe schematic diagram of the multiplexed LC-MRM setup is shown in Fig. 1. A duo channel UPLC system (Thermo Scientific Waltham, MA) equipped with two fast gradient quaternary pumps and two autosamplers was coupled to a QqQ MS (Sciex API 6500, Framingham, MA). Aria software was used to schedule staggered NPLC and HILIC separations. Aria also managed the QqQ MS to acquire only the selected portions of entire HILIC and NPLC separations. A makeup pump was integrated to mix an ionization enhancing solvent (90% IPA containing 10 mM ammonium acetate and 0.1% formic acid) with the NPLC eluent. The separation columns, mobile phases, and representative gradient conditions are shown in supplemental Table S5.

Fig. 1Schematic diagram of the multiplexed NPLC/HILIC-QqQ lipidomics setup. HILIC, hydrophilic interaction chromatography; NPLC, normal phase chromatography; QqQ, triple quadrupole.

The QqQ MS was operated using electrospray ionization with the following parameters: curtain gas 40, ion source gas (1Róg T. Orłowski A. Llorente A. Skotland T. Sylvänne T. Kauhanen D. et al.Interdigitation of long-chain sphingomyelin induces coupling of membrane leaflets in a cholesterol dependent manner., 2Fanning S. Haque A. Imberdis T. Baru V. Barrasa M.I. Nuber S. et al.Lipidomic analysis of α-synuclein neurotoxicity identifies stearoyl CoA desaturase as a target for parkinson treatment.) 80, source temperature 295°C, and collision gas (CAD) 9. Other parameters for positive and negative modes are as follows, respectively: ion spray voltage 5500 V and -4500 V, entrance potential 10V and -10V, and collision cell exit potential 15V and -15V.A library containing thousands of MRM transitions was constructed based on lipid-class specifics as described in supplemental Table S2. Scheduled MRM conditions were optimized based on the retention times (RTs) of synthetic standards. The MRM detection window was set between 40 to 120 s to bracket lipid peaks of the same class. Roughly 2.8 scan cycles per second was used by adjusting target scan time between 0.15 to 0.18 s. The peak integration was performed using MultiQuant (Version 3.03, Sciex). Isotopic correction of the measured lipids was performed using LICAR (21Gao L. Ji S. Burla B. Wenk M.R. Torta F. Cazenave-Gassiot A. LICAR: An application for isotopic correction of targeted lipidomic data acquired with class-based chromatographic separations using multiple reaction monitoring.).Method validationExtraction recovery was based on measuring the difference between pre-extraction and post-extraction spiked deuterated lipid standards normalized to selected endogenous lipids. Extraction recovery was assessed using 46 deuterated synthetic lipid standards (see details in supplemental Table S6).

Three calibration curves were tested in parallel to determine the linearity, dynamic range, limit of quantitation (LOQ), and limit of detection. We defined an acceptable calibration curve with a regression coefficient R2 > 0.99 and minimum six calibrators. We typically define tiered stringencies in precision and accuracy for exploratory measurement versus assays for use within FDA-regulated requirements. Early discovery biomarker analyses require less stringency in terms of acceptance criteria in order to reduce the number of false negative responses in initial analyses prior to follow-up validation. In this case, each calibrator, including limit of quantification, must meet precision (%CV) ≤ 25% and accuracy ± 25% of nominal values. Limit of detection was determined as the lowest standard with a signal-to-noise ratio of at least 3 and twice more than signal-to-noise ratio of single blanks. We use quality control samples (n ≥ 3) either prepared from pooled study samples or biological matrices with similar lipidome compositions in each run to capture the variabilities of all measurable lipid species. Herein, the within-run variabilities were determined using lipid concentrations measured in NIST 1950 plasma undiluted (n=6), 5x and 25x diluted (with 2% BSA, n=6). The experiment was repeated three consecutive times to evaluate between-run reproducibility. The concentrations of all quantified lipids were derived from the calibration curves in each run.

ResultsCombining NPLC and HILIC to extend lipid class separation capacityWe explored the application of multiplexing NPLC-HILIC to improve the separation of lipid classes than the uses of each method alone. Under the HILIC and NPLC chromatographic conditions (supplemental Table S5), a mixture of 26 lipid standards each representing a lipid class was tested on both chromatographic systems. NPLC provided baseline chromatographic separation of neutral lipids including cholesteryl esters (CE), cholesterol, glycerolipids, and free fatty acids and less hydrophilic lipids such as ceramides (Cer) as shown in Fig. 2. Baseline separation was achieved between CE and triacylglycerols (TAG), the two most lipophilic and abundant lipid classes, which is typically limited with HILIC. We obtained separation of diacylglycerols (DAG) 1,2 and 1,3-acyl positional isomers (supplemental Fig. S1A) on NPLC as reported elsewhere (22Hutchins P.M. Barkley R.M. Murphy R.C. Separation of cellular nonpolar neutral lipids by normal-phase chromatography and analysis by electrospray ionization mass spectrometry.). GalCer and GlcCer were not baseline separated by NPLC being reported as hexosylceramide (HexCer). Although HexCer, Cer, lactosylceramide, and sulfatides displayed sharp and symmetrical peak shapes on NPLC, they eluted closely. NPLC offered limited separation to polar lipid classes such as the phospholipids. We decided to divert this part to waste, instead, monitor polar lipids in HILIC mode (Fig. 2). HILIC provided excellent separation for phospholipids classes, furthermore, baseline separation of isomeric classes such as GalCer and GlcCer, and sn-1 and sn-2 lysophospholipid isomers (23Koistinen K.M. Suoniemi M. Simolin H. Ekroos K. Quantitative lysophospholipidomics in human plasma and skin by LC-MS/MS.) (supplemental Fig. S1B). We note that individual lipid species within the same class did not overlap completely, rather eluted with very slightly different RT which could be predicted based on their carbon chain lengths and number of DBs (supplemental Fig. S4). It should be noted that additional lipid subclasses can readily be added into this platform beyond the studied herein. For example, we have observed excellent separation between isomeric phosphatidylglycerols (PGs) and bis(monoacylglycerol)phosphates as reported previously (24Hankin J.A. Murphy R.C. Barkley R.M. Gijón M.A. Ion mobility and tandem mass spectrometry of phosphatidylglycerol and bis(monoacylglycerol)phosphate (BMP).) in other data sets not included in this report. Lysophosphatidic acid was excluded due to peak tailing, which requires alternative chromatographic conditions for improvement.

Fig. 2Chromatographic separation of the sMRM-NPLC/HILIC method providing different levels of quantitative information of measured lipids. Total Ion Chromatograms (TIC) of standards belonging to 26 different lipid classes monitored by NPLC and HILIC are shown respectively. The ion intensities of the measured standards were individually converted to percent distribution (y-axis) to better illustrate the chromatographic separation. The chromatograms were reconstructed in Prism 9.0. The taken chromatographic and MS acquisition strategy offer four levels of quantification, namely class, isomeric, molecular species, and positional isomers, illustrated with TAG, GlcCer, GalCer, and PE. GalCer, galactosyl-ceramide; GlcCer, glucosyl-ceramide; HILIC, hydrophilic interaction chromatography; NPLC, normal phase chromatography; QqQ, triple quadrupole.

MS optimization for accurate lipid identification and quantitative performanceThe high degree of heterogeneity across the molecular structures of glycerolipids, sterol lipids, sphingolipids, and phospholipids introduces distinct challenges to the development of an optimized detection method. We optimized and balanced single-run MS conditions to achieve lipid identification, detection sensitivity, and low in-source fragmentation (ISF). To build MRM transitions, we selected [M+H]+, [M+NH4]+, [M-H]-, and [M+OAc]- precursor ions and monitored multiple structurally characteristic fragment ions of both ion modes (supplemental Table S2). RTs of lipid standards and ISs provide additional important information for identification.Loss in MS detection sensitivity and ISF may occur to some lipids due to improper ion source settings. We observed marked improvements in sensitivity of neutral lipids and sphingolipids using lower ion source temperatures, whereas phospholipids detection was favored at relatively higher temperatures (supplemental Fig. S2). Neutral lipids were observed to be particularly sensitive to temperature-dependent ISF. For example, we observed a strong cholesteryl cation at m/z 369.3 at the RT for CE which resulted from neutral loss of ROOH of ammoniated CEs (25Hutchins P.M. Moore E.E. Murphy R.C. Electrospray MS/MS reveals extensive and nonspecific oxidation of cholesterol esters in human peripheral vascular lesions.). We detected in-source losses of hexose, di-hexose, H2O, and sulfate (SO3) from HexCer, 2HexCer, and sulfatides (supplemental Fig. S3A2) which became more severe at a higher ion source temperature (e.g., 450°C vs. 300°C). ISF is inevitable and brings ambiguity into lipid annotation and quantification when chromatographic separation between classes is insufficient. As a case in point, close evaluation of the NPLC chromatographic data was required to discriminate closely eluting sulfatides, HexCer and dihexosylceramide (Hex2Cer) (Fig. 2). As shown in supplemental Fig. S3C, sulfatide d18:1d7/13:1 was observed to lose SO3 in source even at mild conditions to form an identical ion as GalCer d18:1d7/13:1. This type of ISF artifact can result in a pronounced contamination of the HexCer MRM transitions without sufficient chromatographic resolution. Thus, HILIC is more advantageous than NPLC in this situation due to better separation of these lipid species as shown in supplemental Fig. S3. We settled on 295°C and further softened other ion source parameters (e.g., declustering potential) for lipid classes prone to ISF. Cholesterol and other sterol lipids showed unsatisfactory sensitivity using ESI. A separate LC/MS analysis is required using positive atmospheric pressure chemical ionization detection monitoring [M-H2O+H]+ or [M-2H2O+H]+ (not shown).Multi-fragment MRM for deeper informationWe chose to monitor multiple MS/MS fragments related to the headgroups, FAs, and long-chain bases of the same lipid species to provide additional structural information for peak identification (supplemental Table S2). Although targeted lipidomic methods often do not require multiple fragment ions for quantification, the rationale for adding this additional information in our strategy is two-fold. Firstly, using multiple fragment ions of the same molecular species reduces the false-positive rate (13Liebisch G. Ekroos K. Hermansson M. Ejsing C.S. Reporting of lipidomics data should be standardized., 26Simons B. Kauhanen D. Sylvänne T. Tarasov K. Duchoslav E. Ekroos K. Shotgun Lipidomics by sequential precursor ion fragmentation on a hybrid quadrupole time-of-flight mass spectrometer.) of lipid identification. Mainly, it enables deeper lipidome information. Monitoring both acyl anions of diacylphospholipid species offers a potential to quantify alterations in positional isomers, that is, the FA distribution on the sn-1 and sn-2 positions of the glycerol backbone (27Ekroos K. Ejsing C.S. Bahr U. Karas M. Simons K. Shevchenko A. Charting molecular composition of phosphatidylcholines by fatty acid scanning and ion trap MS3 fragmentation.). To prove feasibility, we analyzed mixtures of isomeric-pure standards of phosphatidylcholine (PC) and phosphatidylethanolamine (PE) of 16:0/18:1 and 18:1/16:0 configurations at varying molar ratios and monitored both acyl anions FAs 16:0 and 18:1 by MRM. By plotting each FA peak area as a percent of both FAs, we obtained a linear response for both PC and PE (supplemental Fig. S5). Thus, from the linear regression lines, we could determine the amount of a particular positional isomer, for example, 30% FA 16:0 (Y-axis, supplemental Fig. S5A) translated to roughly 90% in the form of PE 16:0/18:1 and 10% as PE 18:1/16:0. The current approach is sufficient to estimate the relative change in the distribution of positional isomers, but absolute quantification of each positional isomeric species will require titration curves for each pair of isomeric-pure standards or by acquiring additional structural information by MS3 fragmentation where feasible (