Tools and Techniques for Lipidomics Studies

Lipids are simultaneously the simplest and the most complex biomolecules of interest. While proteins are constructed from combinations of the 21 amino acids and genes from the five nucleotides, the major repeating unit in lipids is -CH_x-. A nearly infinite number of variations exists when branching and unsaturation (with cis- and trans- geometries possible) are included. One could argue that metabolites are more chemically diverse, but that is only because they are solely derived from other molecular classes.

Not just ‘fats’

Say “lipid” and an organic chemist will think of fats, oils, or waxes, but the lipidome is a lot more diverse than those related variations on the hydrocarbon theme. The LIPID MAPS^® Structure Database (LMSD), a public resource for lipidomics, contains 47,449 unique lipid structures, making it the largest lipidomics database in the world. As of April 1, 2022, LMSD incorporates 47,449 unique lipid structures belonging to eight molecular classes: fatty acids, glycerolipids, glycerophospholipids, sphingolipids, sterol lipids, prenol lipids, saccharolipids, and polyketides.

Chain length, positional isomerization, and functionalization are the three obvious structural features, and points of chemical differentiation, within lipid classes. They are also, fortunately, characteristics that are easily distinguishable with modern analysis methods.

Since lipidomics and metabolomics overlap, the former is sometimes referred to as a subset of the latter. While the overlap is significant, the metabolon is, for practical reasons, limited to products of metabolism with molecular weights below around 1500 Da. “Lipids are all hydrophobic, although some classes may be isolated through aqueous extraction,” says Rahul Deshpande, Ph.D., who works on metabolomics application development at Thermo Fisher Scientific. “Because lipids function in unique ways and their analysis involves a limited set of techniques, they are in their own ‘omics category. But keep in mind that some grey area exists.”

Another subtle point sets lipids apart from the other ‘omics. The cellular fractions from a plant or animal sample, when lysed and fractionated for their chemical components, contain proteomic, genomic, and metabolomic products arising from host organism metabolism (both catabolism and anabolism). The lipid fraction, however, may arise from analogous processes in the organism under study, or which are acquired from the diet. “They’re metabolites, but from a different organism than the one under investigation,” Deshpande explains.

Classical lipid analysis involved gas chromatography (GC) of hydrolyzed lipids, with or without derivatization. Today, liquid chromatography-mass spectrometry (LC-MS) is the go-to analytical method, but older techniques are still used, including thin-layer chromatography and GC with flame ionization or mass detection.

Targeted vs. untargeted

Lipidomics studies may be targeted or untargeted. Untargeted studies tend to look at the lipid components of global expression and metabolism, which is where databases are useful. Targeted lipidomics often involves a needle in a haystack approach, where the target may be in very low absolute abundance, or relative to more prevalent species. Targeted lipid analysis may require sample preparation involving enrichment, depletion, or spiking (with or without labeling).

“As with all ‘omics, concentration dynamic range is an issue with lipidomics, especially with targeted experiments, for example on lipid markers of inflammation,” Deshpande tells Biocompare. “There’s also the issue of dynamic or constantly changing concentrations.” Deshpande is referring to biochemical processes that continue after cell lysis that affect lipid concentrations, especially in targeted lipidomics, plus the effects of reagents and environmental exposure (e.g., oxidation). “Even transferring cells to phosphate-buffered saline may affect concentrations.” Membrane lipids are relatively unaffected by these factors, but their composition changes as the cell progresses through its life cycle. “All these factors must be considered when designing experiments.”

The quest for granularity

Although lipids are generally assigned into one of eight categories, these classifications may be further differentiated by chemical structure. “Lipidomics methods are fine-tuned to characterize these structural complexities and are based on the same analytical platforms used for polar metabolite profiling,” says Maryam Goudarzi, Sr. Manager at SCIEX specializing in Life Science Research—Lipidomics and Metabolomics.

These methods begin with cell lysis (where applicable), extraction, and LC (HILIC, reverse phase) with ultraviolet, MS, or MS/MS detection. Within MS, options include MALDI imaging and various ion mobility methods. Gas chromatography is also used for apolar fractions or for derivatized polar moieties, while thin-layer chromatography is useful for screening extracts or chromatography fractions. “Lipidomics may also be paired with any other biochemical or biophysical analysis to answer questions on global metabolism,” Goudarzi says.

As with proteomics, genomics, and metabolomics, lipidomic experiments rely heavily on standard spectral libraries. Lipids are found in most major spectral compendia (NIST, HMDB, mzCloud, etc.), but details on fragmentation and fragmentation conditions differ among these libraries,” Goudarzi explains. “While reporting on lipid classes, carbon number, and unsaturation is routine in lipidomics experiments, complete structural elucidation still proves analytically and computationally challenging. Zeno EAD provides a higher level of confidence in lipid IDs by providing additional fragments that otherwise would be missing in libraries. This is an exciting new opportunity to populate lipid mass spectral libraries with enriched annotations.”

SCIEX has published an application note describing EAD, on the ZenoTOF 7600 MS system, for structural elucidation of glycerophospholipids, sphingolipids, and acylglycerols in a single experiment. Compared with the more commonly used collision induced dissociation, EAD reports on unique fragment ions that are critical for complete lipid characterization.

Key features include a broader, deeper look into the lipid sample, tunable kinetic energies for optimizing fragmentation, and the sensitivity required for fast LC-MS analysis and information-dependent acquisition.