Immunoaffinity Extraction and Bioactive Lipid Profiling

September 15, 2018 (Vol. 38, No. 16)

Cayman Chemical’s Solid-Phase Extraction with LC-MS/MS Improves the Analysis of Trace Lipids

Bioactive signaling molecules derived from the lipid structural components that make up all biological membranes have been known for decades to play critical physiological roles in various organisms and to be involved in human diseases affecting every tissue and organ. Researchers have developed many techniques to determine the levels of lipids in biological samples.

At first, researchers applied classic biological activity assays, such as assays measuring the contraction of guinea pig ileum (the slow-reacting substance of anaphylaxis, later identified as leukotrienes) or the release of serotonin by rabbit platelets (platelet-activating factor, later identified as a phospholipid). For many years, the most widely used quantitative technique for many lipid mediators has been the enzyme-linked immunosorbent assay (ELISA, or EIA), still popular because of its sensitivity, reproducibility, and accessibility to research groups equipped with easy-to-use instrumentation.

More recently, however, researchers and clinicians began turning to mass spectrometry (MS). This analytical technique has the inherent advantage of avoiding any suspected antibody cross-reactivity. In addition, MS offers multiplexing capabilities that overcome the many challenges of quantifying small molecules. Because MS is able to quantify multiple targets from minimal sample amounts with accuracy and certainty, it is becoming increasingly prevalent, even the method of choice.

In this tutorial, we provide an example of how MS may be used to analyze bioactive lipids present in the spleen of mice. This example, which uses what is now a well-established technique, highlights the many advantages of this approach, but also reveals an important pitfall that may be encountered when this approach is applied to certain biological matrices. Finally, this tutorial presents one successful way to sidestep this pitfall, somewhat ironically through the time-honored use of specific antibodies.

The goal of this project was to study the levels of oxylipins, signaling molecules that are mostly derived from arachidonic acid (and include the eicosanoids), in the spleen of healthy mice compared to mice suffering from experimental autoimmune encephalitis.

Isolating Lipids with Solid-Phase Extraction

The initial challenge in lipid analysis is extraction from a biological matrix. Some very widely used protocols for lipid extraction are based on the use of organic solvents. These protocols, such as those originally published by Folch or by Bligh and Dyer, result in almost quantitative extraction of most lipids, including some lipids that are present at very high concentrations in biological tissues (for example, cholesterol, phospholipids, or triglycerides). The prevalent lipids can interfere with the analysis of lipids such as oxylipins, which are usually present at concentrations several orders of magnitude lower.

In this study, we performed solid-phase extraction (SPE) using reversed-phase cartridges known to result in substantial enrichment in free fatty acids, including oxylipins. Since recovery of analytes during extraction can be somewhat variable, it is essential to add an adequate mix of internal standards as soon as extraction begins. We added equal amounts of 12 deuterated internal standards to each sample.

Using the ratio of each analyte to its internal standard minimizes possible differences between extraction efficiencies across samples. Internal standards are also essential to build isotope-dilution calibration curves, allowing for absolute quantitation of analytes. However, we did not use this method here, as only a comparison between experimental groups was required.

Detecting Multiple Lipid Analytes with LC-MS/MS

For relative comparison of the levels of these analytes across samples, we used a combination of liquid chromatography (LC) and tandem MS (MS/MS). LC-MS/MS increases the specificity and sensitivity of detection—the results of which are shown in the Table. From a single experiment, the levels of 20 different lipid analytes could be directly compared across samples. Over 20 additional lipids were detected but not analyzed.

This targeted lipidomic profiling allows quantitation of many lipids over a wide dynamic range in the same sample. Using standard statistical analysis tools, this experiment revealed a changing trend in the 5-lipoxygenase (5-LO) pathway products. Individual differences within each group, as well as differences between the two experimental groups, are visualized with a heat map (Figure 1).

The major products of arachidonic acid oxidation catalyzed by 5-LO constitute a group of lipid mediators of inflammation known as leukotrienes (LTs). Due to their roles in the pathogenesis of diseases like asthma, neurological diseases, or diabetes, LTs are under investigation to further understand their functions in disease progression and to determine their potential value as biomarkers. However, their low abundance and rapid clearance make LTs difficult to measure in biological matrices.

GC-MS has been successfully used to measure LTs, but this technique requires multiple derivatization steps. ELISAs have also been used, but these assays could deliver inaccurate results if antibody cross-reactivity occurs. LC-MS/MS provides outstanding specificity. The technique is based on the three distinct criteria: LC elution time, mass-to-charge (m/z) ratio of the molecular ion, and a characteristic fragment generated upon gas-phase collision within the mass spectrometer. Nonetheless, in LC-MS/MS, unknown molecules sometimes interfere with the analysis.

When an LC-MS/MS-based method was applied to the extract from a healthy mouse brain, a compound showing the same m/z transition as leukotriene C₄ (LTC₄), but eluting shortly before authentic LTC₄, was observed (Figure 2, top middle panel). The signal for LTC₄, if present at all, was buried within a shoulder on the chromatographic peak for this contaminant, which presents a substantial analytical challenge, particularly considering that LTs are present at very low levels in brain tissue.

Table. Relative quantitation by LC-MS/MS of oxylipins in the spleens of mice

Removing Interference with Immunoaffinity Extraction

Trying to overcome this analytical challenge, we performed immunoaffinity extraction (IAE) of leukotrienes, using a mixture of agarose beads bound to antibodies targeted to recognize either LTB₄ or cysteinyl-leukotrienes (LTC₄, LTD₄, and LTE₄). As shown on the top left panel of Figure 2, this resulted in a clean signal for LTC₄ at the correct elution time, with little if any interference by the contaminant observed in the SPE samples. These results validate IAE as an alternative to SPE and indicate that IAE can handle biological matrices that may include interfering compounds.

An obvious drawback of this approach is that it allows the analysis of only those lipids recognized by the available immobilized antibodies. However, as shown in the middle panels of Figure 2 (these panels show PGE₂ as an example), it is possible to obtain a full eicosanoid profile from a challenging matrix by performing IAE and using its flow-through fluid to perform SPE. The use of internal standards minimizes the problem of diminishing recoveries when this double extraction protocol is used.

Conclusion

In summary, we present a method for the LC-MS/MS-based analysis of bioactive lipids in two different biological matrices, including one that presents the challenge of a closely eluting contaminant. This method has multiple benefits over traditional SPE:

1. The specificity of the antibody provides minimal risk of column overloading, which allows for the use of large sample volumes if necessary.
2. Samples are cleaner, minimizing problems associated with matrix interferences.
3. The promiscuous nature of antibodies, a critical concern about their use in immunoassays, is overcome by the specificity of LC-MS/MS.
4. Where necessary, the flow-through fluid can be used to analyze compounds not retained by the antibodies used.

Overall, this report verifies the usefulness of SPE combined with LC-MS/MS to analyze lipid mediators in biological samples. Furthermore, it offers the possibility of using IAE enrichment to improve the analysis of some lipids present in trace amounts in certain challenging samples, overcoming some limitations of both immunoassays and MS, and adding a useful tool to the existing methods available for research and clinical laboratories.

Figure 2. LC-MS/MS chromatograms of LTC4 (top panels) and PGE2/PGD2 (middle panels) from the brain of a healthy mouse. Both PGE2 and PGD2 exhibit the m/z 351-to-271 collision-induced transition, but they can be resolved chromatographically and identified by coelution with deuterated internal standards (not shown). The bottom panels summarize the extraction strategies used to obtain the data shown above them.

Miguel A. Gijon, Ph.D., is an analytical chemist and Paul D. Kennedy, Ph.D., is director of analytical chemistry at Cayman Chemical. Athena M. Soulika, Ph.D., assistant professor of dermatology at the University of California, Davis, provided cell and tissue samples. E-mail: [email protected].