Lipidomics workflows – Why lipidomics?

Why lipidomics?

Lipidomics is a subset of metabolomics that has evolved into a class all its own. Lipidomics is the attempt to map and quantify lipid species sets within a cell or tissue to identify biomarkers and elucidate metabolism at the cellular level.

Lipidomics workflows are diverse and complex; however, the development of HRAM mass spectrometry has risen to the challenge of lipid analysis and offers an unprecedented level of sensitivity, selectivity and precision.

Receive updates on Metabolomics  Join community

Metabolomics Software Solutions

This lipidomics workflows overview reviews the challenges of lipidomics and the strategies employed for analysis.


Lipid classification

Lipids are heterogeneous compounds that are insoluble in water, soluble in organic solvents and chemically and structurally diverse. The number of distinct lipid chemical species is estimated to be between 104–105.

The LipidMaps consortium has developed the first internationally accepted lipid classification system1,2, as well as a nomenclature and structural representation system. There are eight lipid categories, each complete with its own classification and sub-classification hierarchy.

CategoryAbbrev.Example
Fatty acylsFAoleic acid (1-octadecenoic acid)
GlycerolipidsGL1-hexadecanoyl-2-(9Z-octadecenoyl)-sn-glycerol
Glycero-phospholipidsGP1-hexadecanoyl-2-(9Z-octadecenoyl)-sn-glycero-3-phosphocholine
SphingolipidsSPN-(tetradecanoyl)-sphing-4-enine
Sterol lipidsSTcholest-5-en-3β-ol
Prenol lipidsPR2E,6E-farnesol
SaccharolipidsSLUDP-3-O-(3R-hydroxy-tetradecanoyl)-αd-N-acetylglucosamine
PolyketidesPKaflatoxin B1

Lipidomics and mass spectrometry

Due to the extensive chemical diversity of lipid species, it is a challenge to measure the lipidome comprehensively and in a single experiment. The following are just a few of the difficulties inherent in lipidomics analysis, and also why these challenges require highly sensitive and fast scan LC-MS systems.

Lipidomics challengesRequirements of LC-MSn platform
Diversity in structures and physical chemical propertiesHigher resolving power for both the HPLC separation and for the MS
Thousands of isomeric and isobaric speciesFaster effective MS/MS scan speed and excellent mass accuracy
Very low to very high concentrationsHigh sensitivity and wide dynamic range for both MS and MS-MS
Time consuming lipid IDDedicated software for automated lipid molecular ID and quantification

Untargeted lipidomics often begins with exploratory experiments, which are then followed by targeted analysis of specific pathways.

High resolution mass spectrometry approaches have been successful in quantitative profiling, lipid identification and some structural elucidation; however, other methods such as chemical derivatization or 2D NMR may be needed for the unequivocal assignment of structures such as double bond configurations.

References

  •  Fahy, E et al.(2005) A comprehensive classification system for lipids. J Lipid Res 46: 839-861. PubMed
  •  Fahy, E et al. (2009) Update of the LIPID MAPS comprehensive classification system for lipids. J Lipid Res 50: S9-S14. PubMed

Untargeted or discovery lipidomics: shotgun workflow

The infusion or shotgun lipidomics workflow does not rely on a chromatographic separation; instead, it relies on direct infusion of the sample into the mass spectrometer. Because there is no chromatography time scale involved with the eluting peaks, the molecular ions of individual molecular species in a lipid class remain at constant concentration during mass spectrometry data acquisition. This means that several MS scan functions can be performed for both identification and quantitation without the time constraint typically encountered with “on the fly” analysis during chromatographic elution.

The advantage of the shotgun workflow is the speed at which one can rapidly profile large numbers of samples, defining which lipid classes are significantly changing prior to conducting a more in-depth analysis. One drawback of the shotgun lipidomics workflow is that many isobaric species may overlap, and especially in a crude lipid. Therefore, high mass resolution and/or multiple stages of MSn and dissociation are required for confident identification and quantitation.

One example of the shotgun workflow is demonstrated below. This workflow uses a very high resolution mass spectrometer and flexible fragmentation modes that accurately distinguish isobaric lipid species. Isotopic fidelity, a dynamic quantitation range, and HCD and CID for in-depth fragmentation analysis are also employed. This workflow was chosen due to its sensitivity and the high throughput needs of the comparative study of mouse cerebellum and hippocampus.

This lipidomics study reproducibly quantified 311 lipid species spanning over 20 lipid classes; it also resulted in the identification of 202 distinct molecular glycerophospholipid species.

References

  •  Almeida R, Pauling JK, Sokol E et al (2015) Comprehensive Lipidome Analysis by Shotgun Lipidomics on a Hybrid Quadrupole-Orbitrap-Linear Ion Trap Mass Spectrometer. Eur J Lipid Sci Technol 117(10): 1540–1549. SpringerLink
  •  Surma MA, Herzog R, Vasilj A et al (2005) An automated shotgun lipidomics platform for high throughput, comprehensive, and quantitative analysis of blood plasma intact lipids 26(1): 133–148. PMC
  •  Ryan E and Reid GE (2016) Chemical Derivatization and Ultrahigh Resolution and Accurate Mass Spectrometry Strategies for “Shotgun” Lipidome Analysis. Acc Chem Res 49(9): 1596-1604. ACS
  •  Milgin M, Born P, Fezza F et al (2016) Lipid Discovery by Combinatorial Screening and Untargeted LC-MS/MS. Sci Rep 6: 27920. PMC

Untargeted or discovery lipidomics: LC-MS workflow

Another untargeted workflow, which is also called discovery lipidomics, involves using chromatographic separations (HPLC) connected to mass spectrometry to compare the lipidome between control and test groups, and to identify those differences between lipid profiles that may be relevant to specific biological conditions. This workflow is very similar to a typical proteomics workflow.

It is critical that the lipidomics LC-MSn platform offer high resolving power on both HPLC separation and MS detection to separate and distinguish the many isobaric and isomeric species found within biological lipid extracts. When compared to direct infusion methods, HPLC adds retention time as another dimension to its selectivity, including separation of isomeric species that have the same exact mass and molecular formulae or “sum composition”. Lipid separations can be optimized using various columns, such as the C30 and C18 reversed phase columns.

Lipidomics search process – MS data › Peak detection

Following data acquisition, peak intensities of both the precursor molecule and its MS2 fragments are calculated.

Together with HRAM LC-MS systems, the combination of HPLC and LC-MS in lipidomics delivers retention time, exact ppm precursor masses, and product ion spectra as different means of identification. High mass accuracy paired with retention time is particularly useful for lipid identification because there may be too little information presented in the often sparse fragmentation patterns of lipid MS/MS spectra.

There are three steps in this workflow:

Identification and quantitation

Identification is performed using both the MS and MS/MS spectra, with quantitation occurring only at the MS level of the profiling experiment.

Lipidomics search process – Identification › Quantitation

Lipid identification

After peak detection, data-dependent MS2 files are matched against lipid databases to predict known fragmentations from reference compounds. The available software should perform the following tasks:

  • Support the appropriate mass spectrometry data; ie., high resolution accurate mass versus nominal mass.
  • Have the ability to peak detect and search LC data-dependent raw files.
  • Predict MS2/MS3 spectra of precursor ions.
  • Align data annotations within a user-defined retention window.
  • Merge positive and negative ion data into one view.
  • Support the combination of HCD and CID fragmentation data into one search.

Lipid quantitation

Quantitation is performed using the extracted ion chromatograms (XIC) of precursor masses:

Statistical analysis

  • Lipidomics samples are typically complex and there are multiple interactions between metabolites and lipids in biological states. To uncover significant events, univariate and multivariate statistical analysis (chemometric methods) platforms use visualization tools to assess abundance relationships between different lipid components.
    1. Univariate methods are the most common statistical approach and analyze lipid features independently. When assessing differences between two or more groups, parametric tests such as student’s t-test and ANOVA (analysis of variance) are commonly used.
    2. Multivariate methods analyze lipid features simultaneously and can identify relationship patterns between them. Principal component analysis (PCA) is a common example of a multivariate method approach.

Interpretation

There are several ways to interpret data once all lipids have been identified. Data interpretation depends on initial experiment design, putative biomarker discovery, and fingerprinting or mapping pathways in order to understand metabolism.

The biological knowledge available from lipidomics studies has been continuously and rapidly increasing, and groups of lipids related to the same biological process have been mapped to metabolic pathways. However, there is a lack of biological databases containing information on relevant lipid molecular structures as related to specific pathways.


Structural elucidation of lipids

For ultimate profiling and identification where typical high resolution and MS2 are not sufficient, advanced mass spectrometry with ultra high resolution (500k FWHM), MSn capabilities and several dissociation modes offers more confident identifications and higher throughput. A new class of mass spectrometer, called the Tribrid (because of its three mass analyzers) allows for the following types of experiments:

Structural elucidation of lipids – Tribrid

Because untargeted and targeted identification can occur in the same experimental workflow, this allows the researcher to design experiments that answer specific lipid structural questions while LC/MS and data dependent-MS2 are being performed. The advantage of such a process is that you can design a targeted experiment to answer a specific structural question while performing a data-dependent profiling experiment.

Using high resolution MS analysis and HCD-triggered CID MS2, you may identify PC isomers from the positive ion fragment ions:


Identification and quantitation

Identification is performed using both the MS and MS/MS spectra, with quantitation occurring only at the MS level of the profiling experiment.

Lipidomics search process – Identification › Quantitation

Lipid identification

After peak detection, data-dependent MS2 files are matched against lipid databases to predict known fragmentations from reference compounds. The available software should perform the following tasks:

  • Support the appropriate mass spectrometry data; ie., high resolution accurate mass versus nominal mass.
  • Have the ability to peak detect and search LC data-dependent raw files.
  • Predict MS2/MS3 spectra of precursor ions.
  • Align data annotations within a user-defined retention window.
  • Merge positive and negative ion data into one view.
  • Support the combination of HCD and CID fragmentation data into one search.

Lipid quantitation

Quantitation is performed using the extracted ion chromatograms (XIC) of precursor masses:

Statistical analysis

  • Lipidomics samples are typically complex and there are multiple interactions between metabolites and lipids in biological states. To uncover significant events, univariate and multivariate statistical analysis (chemometric methods) platforms use visualization tools to assess abundance relationships between different lipid components.
    1. Univariate methods are the most common statistical approach and analyze lipid features independently. When assessing differences between two or more groups, parametric tests such as student’s t-test and ANOVA (analysis of variance) are commonly used.
    2. Multivariate methods analyze lipid features simultaneously and can identify relationship patterns between them. Principal component analysis (PCA) is a common example of a multivariate method approach.

Interpretation

There are several ways to interpret data once all lipids have been identified. Data interpretation depends on initial experiment design, putative biomarker discovery, and fingerprinting or mapping pathways in order to understand metabolism.

The biological knowledge available from lipidomics studies has been continuously and rapidly increasing, and groups of lipids related to the same biological process have been mapped to metabolic pathways. However, there is a lack of biological databases containing information on relevant lipid molecular structures as related to specific pathways.


Structural elucidation of lipids

For ultimate profiling and identification where typical high resolution and MS2 are not sufficient, advanced mass spectrometry with ultra high resolution (500k FWHM), MSn capabilities and several dissociation modes offers more confident identifications and higher throughput. A new class of mass spectrometer, called the Tribrid (because of its three mass analyzers) allows for the following types of experiments:

Structural elucidation of lipids – Tribrid

Because untargeted and targeted identification can occur in the same experimental workflow, this allows the researcher to design experiments that answer specific lipid structural questions while LC/MS and data dependent-MS2 are being performed. The advantage of such a process is that you can design a targeted experiment to answer a specific structural question while performing a data-dependent profiling experiment.

Using high resolution MS analysis and HCD-triggered CID MS2, you may identify PC isomers from the positive ion fragment ions:



Targeted lipidomics

Targeted lipidomics is quantitative and takes information from discovery experiments or from the literature and/or clinical observations to test the model or the hypothesis. This involves verification and validation of defined groups of known lipids across large sample sets. These experiments require accuracy, high throughput and reliability. For large scale targeted profiling, an HRAM MS/MS workflow is the optimal method; for routine quantitation, SRM on a triple quadrupole MS is the preferred solution.

Large scale targeted profiling

Parallel reaction monitoring (PRM) offers a new perspective for targeted quantitation compared with the gold standard of quantitation using SRM and triple quadrupole MS. Using HRAM mass spectrometry, PRM enables greater selectivity, confirmation of data with MS/MS, and retroactive data analysis. As a result, there is a reduced need for method development.

Targeted workflow – PRM acquisition

In the targeted lipidomics profiling workflow, analytical or technical reproducibility is the key to experiment success. High analytical reproducibility means that findings are due to biological variance; it also means that a lower number of samples must be run because technical replicates are minimized.

The three steps in this targeted profiling workflow include:

  • Lipid candidate list creation and method optimization. Ideal fragment ions are selected from discovery experiments, collision energy values are adjusted, and retention times are confirmed.
  • PRM acquisition (link to section on data analysis page). Multiple simultaneous fragmentations are generated and analyzed.
  • Data analysis. Quantitation is performed using the XIC for each metabolite. In experiments where there are samples from test and experimental groups, statistical analysis tools are used for biological insight. As high resolution MS/MS data are collected, and based on the retention time of the accurate m/z precursor, the identity of the lipids is confirmed. Software such as LipidSearch can match the accurate mass MS2 data to the predicted fragment ions for the targeted precursor.

Routine targeted quantitation

For routine quantitation of well-characterized lipids or low-level “signaling” lipids, selective reaction monitoring (SRM) using triple quadrupole MS is the gold standard. The advantages of SRM include reduced interference, lower detection levels and faster method creation and data acquisition. SRM transitions are developed for targeted lipids to produce the best signals.

SRM-based method development is more time consuming compared to the HRAM quantitation approach. Lipids are notorious for having carryover and may also have issues with specificity depending on their isobaric and isomeric overlaps. SRM approaches are optimal for lower-level species with good chromatography data and sample preparation technique for reduced complexity.

Targeted workflow – SRM acquisition

The three steps in the targeted routine quantitation workflow include:

  • Lipid candidate list creation and method optimization. For method optimization, each lipid sub-class requires determination of its unique diagnostic fragment ions. Optimal MS collision energy, authentic standards and a quantitation calibration method are all required for accurate quantitation.
  • SRM acquisition. Batches of samples are tested using the optimized method and appropriate controls, standards and blanks.
  • Data analysis. Using quantitation and reporting software, sample analysis is performed and calibration curves, ion ratios and peak integrations are monitored. Typically, this process is interactive, so samples that do not match a certain user criterion can be flagged.  
Routine Targeted Metabolomics Data Analysis

Resources

Share
 
Style Sheet for Global Design System
CMD SchemaApp code