Robert C. Murphy¹ and Miguel Gijón²

¹University of Colorado Anschutz Medical Campus, Aurora, CO; ²Cayman Chemical Company, Ann Arbor, MI

A Brief History of Quantifying Bioactive Lipids

Over the past 50 years, mass spectrometry has emerged as the gold standard for the quantitative analysis of bioactive lipids, including metabolites of arachidonic acid. Significant advances in mass analyzer hardware and ionization processes have brought this technique to the forefront. The first quantitative studies of prostaglandins involved the use of a magnetic sector mass spectrometer or a single quadrupole mass analyzer coupled to a gas chromatograph, recording a small number of specific mass-to-charge ratios (m/z). In order to carry out these experiments, it was necessary to derivatize the prostaglandin to make it sufficiently volatile to pass through the gas chromatograph. Electron ionization imparted a large amount of internal energy to neutral molecules, resulting in intramolecular decomposition reactions and production of fragment ions, which were then monitored to specifically measure target molecules. The important principle realized from these early experiments was that the mass spectrometer was a remarkable quantitative tool when measuring ion abundances, but only if taking exquisite care of the experimental details. Specifically, the importance of an internal standard was immediately recognized.^1,2 The internal standard corrected for the large number of instrumental variables that determined the mass spectrometer response and thus, the correlation between concentration of a molecular target and abundance of its fragment ions. The internal standard also helped identify analytes by chromatographic co-elution. Although various quantitation strategies could be employed, the most widely used were based on calibration curves to establish the relationship between the concentration of a target analyte and the measured ion abundance relative to that of the internal standard when prepared under identical isolation and derivatization steps.

Important changes have occurred just in the past 25 years, with the emergence of electrospray ionization (ESI) and remarkable developments in tandem mass spectrometry, including tandem quadrupole mass spectrometer systems and ion trap-based mass analyzers. ESI has eliminated the need to derivatize nonvolatile molecules, such as eicosanoids, while generating abundant carboxylate anions from most of these bioactive lipids. However, ESI is a rather low-energy process, rarely yielding useful fragment ions. Both in tandem quadrupole and in ion trap mass spectrometers, this limitation is overcome through collision of ions with neutral gas molecules in order to increase the internal energy of precursor ions and initiate intramolecular rearrangement and fragmentation processes, resulting in product ions. The advantage of tandem mass spectrometry is to allow precise measurement of precursor-product ion relationships, adding important molecular signatures that strengthen the identification of the molecule. Another important advancement has been the ability to drastically increase the number of precursor-product ion pair transitions that can be monitored during a single duty cycle (i.e., the amount of time taken to monitor one complete series of transitions, repeated during the entire chromatographic run). At first, it was possible only to measure five to ten ion pairs in a duty cycle, but improvements in ion detector technology and fast-scanning electronic circuits have made it possible to measure tens to hundreds of ion transitions.

Stable Isotope-Labeled Internal Standards

The availability of stable isotope-labeled internal standards for eicosanoids has developed hand-in-hand with the advances in mass spectrometer instrumentation and ionization processes. Isotopic variants differ only in their molecular weight from unlabeled target eicosanoids. Such molecules behave in an identical manner to the naturally occurring molecules during sample preparation from the biological matrix. Any potential loss of eicosanoid is completely compensated by the loss of the stable isotope-labeled internal standard. A critical point is to add the internal standard as soon as possible to the sample, so that the ratio of endogenous eicosanoid to internal standard is established before any potential physical loss or chemical degradation. Once this ratio is set, it corrects for any problems associated with isolation and chromatography.

The behavior of the stable isotope-labeled eicosanoid is identical to that of the unlabeled natural product with the exception of kinetic isotope effects. These effects are typically quite small even for deuterium-labeled analogs but become observable during separation by HPLC or capillary gas chromatography. Usually, the deuterated analog precedes elution of the unlabeled species, most likely due to tighter carbon-deuterium bonds relative to carbon-hydrogen bonds, making the molecule somewhat smaller. Separation depends on the number of deuterium atoms in the standard, and one observes no effect in the chromatography of standards labeled with carbon-13 or oxygen-18.

A second important aspect of labeled internal standards is their behavior upon mass spectrometry analysis, in particular collision-induced dissociation. The m/z of the molecular ion will always be shifted by the excess mass of labeled atoms in the molecule, but the product ion may or may not be, depending upon the ion chemistry leading to its formation and the exact positions of the stable isotopes in the molecule. This behavior is also quite useful in trying to understand the mechanism by which ions arise in the collisional activation process.³

It is important to consider the total number of stable isotopes present in the molecule because of the occurrence of natural stable isotopes, in particular carbon-13, and to experimentally determine the population of isotope-labeled species in internal standards by calculating the atom percent excess over the natural abundance of carbon-13. While it is essential to know precisely the m/z of internal standard molecular and fragment ions in a quantitative assay, one cannot assume that, for example, PGE₂-d₄ is 100% d₄ with no d₃, d₂, d₁, or d₀ variants.

The abundance of the d₀ variant (i.e., the unlabeled eicosanoid) is of great concern because it has an important influence on the standard curve generated. Standard curves are prepared by adding increasing amounts of reference standard to fixed amounts of labeled internal standard. We will use a theoretical example in which we add either LTB₄-d₁ (containing 10% LTB₄-d₀) or two different LTB₄-d₄ internal standard preparations (containing either 10% or 0.1% LTB₄-d₀) to a final 1 pM concentration in methanol/ water, then add increasing concentrations (0.1 fM to 30 nM) of unlabeled LTB₄ (Figure 1). After injecting into a reversed-phase LC-MS/MS system, the abundance of ion transitions from m/z 335 (LTB₄-d₀) to m/z 196 or the abundance of ion transitions from m/z 339 (LTB₄-d₄) to m/z 197 are determined (assuming [5-d₁]LTB₄ and [6,7,14,15-d₄]LTB₄ internal standards, respectively). In this experiment, the three different internal standards illustrate the influence that isotope content has on the calibration curve dynamic range and the asymptotic lines where the ratio of LTB₄ to internal standard becomes constant. In order to carry out the quantitation, it is necessary to take into account the deuterated isotope content of the three different labeled LTB₄ preparations, as well as the naturally occurring carbon-13 content. As seen in Figure 1A, when 10% LTB₄-d₀ (m/z 335) is in the LTB₄-d₁ internal standard, the standard curve has a very narrow dynamic range in that it is linear only over a very small portion of ratios m/z 335/336. The useful range is only 0.3 to 10 pM. Dynamic range is considerably increased when using LTB₄-d₄, even with 10% LTB₄-d₀ present (Figure 1B). Specifically, this is because the probability of LTB₄ containing four carbon-13 atoms is very low. The useful range of this standard curve is 0.03 to 100 pM. The theoretical ratio of analyte-to-internal standard becomes constant at 0.1 pM, at the lower end of the curve for both internal standards. When the content of LTB₄-d₀ is reduced to 0.1%, a much wider dynamic range is observed from 0.03 to 100 pM (Figure 1C). Notice that the intermediately labeled species (e.g., LTB₄-d₃ at 40% in this example) has no effect on the standard curve whatsoever. In general, the overall amount of internal standard added to samples is rather immaterial as long as it is held constant. Precise addition is critical, though.

Figure 1. Theoretical standard curves for the quantitation of LTB₄ adding equal amounts (1 pM) of three different deuterium-labeled internal standards to various quantities of LTB₄ in solution (0.0001 to 30,000 pM). The signals from LTB₄ using negative-ion electrospray mass spectrometry were measured from m/z 335 to m/z 339. A. Internal standard having only one deuterium atom and containing 10% unlabeled LTB₄. B. Internal standard having four deuterium atoms and containing 10% unlabeled LTB₄. C. Internal standard having four deuterium atoms and containing 0.1% unlabeled LTB₄. The tables under each graph show the measured isotopic abundance of the different ions relative to the molecular ion of the internal standard (arrows).

In general, standard curves using a stable isotope-labeled internal standard are sigmoid, with two asymptotic regions. The region at the left is driven by the isotopic purity of the internal standard at the molecular weight of the eicosanoid, while the right side of the asymptotic region corresponds to the natural abundance of carbon-13 and oxygen-18 at the molecular weight of the labeled standard (Figure 2). Since standard curves are most often employed for small amounts of eicosanoids, the total number of atoms increasing the mass of the internal standard becomes most important. Thus, to achieve maximum dynamic range of an analytical assay it is critical to optimize hydrogen-deuterium exchange chemistry or incorporation of oxygen-18 in the carboxylate moiety of standards by either chemical or enzymatic means4 to minimize the presence of unlabeled material.

Figure 2. Theoretical standard curve for stable isotope dilution of LTB₄ over a large dynamic range, indicating the asymptotic regions at the limits of the assay which are influenced by the total mass shift (number of stable isotopes) of the internal standard (IS) and the isotopic purity.

Reference Standards

Often overlooked is the importance of reference standards in performing quantitative assays. The purity of the reference standard establishes the accuracy of the method. Of course, if the reference standard is a sodium or ammonium salt of the carboxylate anion, it needs to be taken into account when calculating the molarity of the standard dilutions, but the presence of unknown impurities will also lead to errors in accuracy. Oftentimes, different levels of eicosanoids are reported between laboratories, in part because of varying purities of the reference standards used. Unfortunately, there is no convenient way to assess the quantity of a prostaglandin or an unsaturated fatty acid except for gravimetric measurements. This is not the case for leukotrienes or HETEs, where one can employ Beer’s law to calculate the concentration of a solution based upon UV absorption and known molar extinction coefficients of these conjugated olefins. Unfortunately, there is no convenient way to assess the quantity of a prostaglandin or an unsaturated fatty acid except for gravimetric measurements. This is not the case for leukotrienes or HETEs, where one can employ Beer’s law to calculate the concentration of a solution based upon UV absorption and known molar extinction coefficients of these conjugated olefins.

It is also important to carefully store any reference standard solutions, since instability of any of the eicosanoids, frequently due to oxidation, would lead to errors when generating the standard curve. This is not as critical for stable isotope-labeled solutions. As long as one prepares the standard curve for each batch of analyses, the quantity of isotope-labeled internal standard added to each sample is invariant, even though it may not be accurately known.

Of course, it is still essential to ensure that no oxidation or degradation products of these standards interfere with any of the ion transitions being measured.

In summary, continuing advances in mass spectrometry instruments and increasing availability of high-quality analytical standards are allowing the accurate quantitation of eicosanoids by scientists around the world. As new lipid mediators keep being discovered, it is an ongoing challenge to maintain the availability of adequate tools for the study of these molecules and their roles in physiology and disease.

Article References

1. Sweeley, C.C., Elliott, W.H., Fries, I., et al. Mass spectrometric determination of unresolved components in gas chromatographic effluents. Anal. Chem.38(11), 1549-1553 (1966).

2. Samuelsson, B., Hamberg, M., and Sweeley, C.C. Quantitative gas chromatography of prostaglandin E₁ at the nanogram level: Use of deuterated carrier and multiple-ion analyzer. Anal. Biochem. 38(1), 301-304 (1970).

3. Murphy, R.C., Barkley, R.M., Zemski Berry, K., et al. Electrospray ionization and tandem mass spectrometry of eicosanoids. Anal. Biochem. 346(1), 1-42 (2005).

4. Murphy, R.C. and Clay, K.L. Preparation of ¹⁸O derivatives of eicosanoids for GC-MS quantitative analysis. Methods Enzymol. 86, 547-551 (1982).

About the Authors

Robert C. Murphy, Ph.D.

Dr. Murphy is an Emeritus Professor in the Department of Pharmacology at the University of Colorado Anschutz Medical Campus. He has dedicated much of his life to the study of bioactive lipids, largely using and developing sophisticated mass spectrometry techniques. By elucidating the structure of Slow-Reacting Substance of Anaphylaxis (SRS-A), which he termed leukotriene C₄, he forged novel avenues for research on a unique pathway of arachidonic acid metabolism. He has mentored many scientists and influenced careers. He has received numerous awards throughout his own career, including his election as President of the American Society for Mass Spectrometry and serving on its Board of Directors.

Miguel Gijón, Ph.D.

Dr. Gijón is a scientist at Cayman Chemical. His career interests, sparked by the study of lipid mediators of inflammation, include the biological roles of lipids in disease, the catalytic mechanisms and regulation of enzymes implicated in lipid metabolism, and the detailed description of lipid composition in cells and tissues. He is currently a key member of the lipidomics services team, developing or adapting lipid extraction and mass spectrometry-based analysis protocols, as well as discussing experimental models with other researchers to find the most useful approaches to their lipid analysis needs. He maintains active collaborations with academic scientists.