Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Jun 1.
Published in final edited form as: Free Radic Biol Med. 2013 Mar 14;59:2–13. doi: 10.1016/j.freeradbiomed.2013.03.001

Quantification of Fatty Acid Oxidation Products Using On-line High Performance Liquid Chromatography Tandem Mass Spectrometry

Bruce S Levison a, Renliang Zhang a,b, Zeneng Wang a, Xiaoming Fu a, Joseph A DiDonato a, Stanley L Hazen a,b,c
PMCID: PMC3772641  NIHMSID: NIHMS454369  PMID: 23499838

Abstract

Oxidized fatty acids formed via lipid peroxidation are implicated in pathological processes such as inflammation and atherosclerosis. A number of methods may be used to detect specific oxidized fatty acids containing a single or multiple combinations of epoxide, hydroxyl, ketone and hydroperoxide moieties on varying carbon chain lengths from C8 up to C30. Some of these methods are nonspecific and their use in biological systems is fraught with difficulty. Measures of specific-oxidized fatty acid derivatives help in identifying oxidation pathways in pathological processes. We used liquid chromatography coupled with electrospray ionization tandem mass spectrometry (LC-MS/MS) as efficient, selective and sensitive methods for identifying and analyzing multiple specific fatty acid peroxidation products in human plasma and other biological matrices. We then distilled the essential components of a number of these analyses to provide an efficient protocol by which fatty acid oxidation products and their parent compounds can be determined. In this protocol, addition of synthetic internal standard to the sample, followed by base hydrolysis at elevated temperature, and liquid-liquid phase sample extraction with lighter than water solvents facilitates isolation of the oxidized fatty acid species. These species can be identified and accurately quantified using stable isotope dilution and multiple reaction monitoring. Use of a coupled multiplexed gradient HPLC system on the front end enables high-throughput chromatography and more efficient use of mass spectrometer time.

Keywords: oxidized fatty acids, eicosanoids, HETEs, HODEs, electrospray ionization, Prostaglandin F2α, lipid peroxidation, mass spectrometry, column switching, multiplexed HPLC

INTRODUCTION

Assessment of lipid oxidation levels is essential to the study of biologically relevant reactive oxygen species[1]. Numerous methods exist to ascertain these levels through the systematic analysis of the oxidized fatty acids that originate from the oxidized lipids both in-vivo and in-vitro[2–25]. Aside from exploring the bulk oxidation, many of these oxidized lipids and their associated oxidized fatty acids have activities of significant clinical interest[3–6, 10, 14, 18, 20, 22, 26–31]. Numerous GC, LC[32], and hybrid methodologies like GC/MS[3, 5, 14, 18, 20, 22, 31] and LC/MS[2–4, 8–10, 12, 19, 23, 27, 33] exist to quantify the levels of these compounds. These markers can be used clinically to ascertain the effectiveness of antioxidant therapies; however, there are no universal guidelines regarding the scale of the analyses required for the determination of lipid oxidation (and hence fatty acid oxidation) products in a large clinical study. We have performed and optimized these types of analyses for nearly two decades in our laboratory and are presenting our insights into what makes for a good high-throughput oxidized lipid assay.

PRINCIPLES

Liquid/Liquid Extraction of Fatty Acids and their Oxidation Products

Recovery of free fatty acids and their oxidation products using the classic Folch[34] and Bligh and Dyer[35] extractions are not easily adapted to the small sample volumes required for high-throughput clinical analyses. Hara's[36] procedure using hexane and isopropanol can be modified and adapted to these small sample volumes (see protocol). With this method, aside from avoiding the toxicity and regulatory issues associated with chlorinated solvents, the organic layer floats on top and is more easily removed. Hexane being a lighter than water solvent allows one to avoid pushing a pipette or needle through the aqueous layer and risking possible contamination that can occur with heavier than water chlorinated solvents. The extraction is efficient since the initial solution is homogenous giving a layer that fully permeates and extracts the entirety of the sample solution. In the next step more hexane is added to generate a heterogeneous mixture from which the top layer is easily removed after centrifugation. Even when this liquid-liquid extraction is done manually, its simplicity, low cost, and efficiency allows for rapid processing of multiple samples simultaneously in a short period of time.

Use of Heavy Isotope Standards

Internal standards should be incorporated before the extraction step to control for any losses that may occur through poor extraction, inefficient removal of the organic phase or adherence to the container walls, etc. Additionally, one may incorporate a distinct stable isotope labeled internal standard of the parent fatty acid to monitor and control for any artificial oxidation products that form during lipid extraction[2, 8, 9, 12, 19, 23, 25, 37]. Preferably these are universally labeled 13C unsaturated fatty acid derivatives; however, the deuterated analogs are generally less expensive and more readily available from commercial sources.

Advantages of Column Switching

For analyses of biological matrices composed of multiple distinct fatty acid species (both oxidized and un-oxidized), a broad gradient elution is often useful to allow for chromatographic separation among the many different compounds present. Once the gradient has been run it is necessary to re-equilibrate the column back to its initial state, which takes 10 to 15 minutes. This re-equilibration period is followed by the injection of the next sample, which with an autosampler could take several minutes followed by one to two minutes (preferred- after the solvent peak) until any peak of interest elutes. During this period the mass spectrometer is sitting idle and not collecting any relevant data from analyte peaks. To make optimal use of the mass spectrometer, the most costly piece in the instrument string, a second column from another HPLC system can be brought on-line that has already been injected with another sample immediately at the end of the first column's run[28, 33].

Multiple-reaction monitoring (MRM) analysis provides an efficient method for quantification of free fatty acids and their oxidized derivatives

Due to the higher specificity and sensitivity of the MRM mode of operation, it was used for the quantification of oxidized free fatty acids and their precursors. Full ion scan ESI/MS in negative ion mode was used for identification of the specific mass-to-charge ratio (m/z) for each of the fatty acid species monitored. A product ion scan (MS2) was used for identification of the corresponding product ions for each of the analytes. The specific precursor and product ions for each of the oxidized arachidonic acids including HETEs, EETs, oxoEETs and PGF2α are shown in Fig. 1. The structure and precursor product ion transitions for oxidized linoleic acids including HODEs and oxoODEs are shown in Fig. 2. The product (daughter) ions chosen for the MRM transitions were selected for their specificity for the distinct structural isomers, allowing quantification of the individual isomers even if they co-elute during chromatography. The MRM mode of operation shows improved performance over selected-ion mode, particularly in cases of high background noise and those with overlapping peak separation (poor chromatographic resolution) of the isomers.

Fig. 1. Precursor ion →product ion transitions and suggested structures of the major product ion derived from each eicosanoid.

Fig. 1

Open in a new tab

ESI negative-ion full scan was used for identification of precursor ions for each of the analytes and ESI negative-ion product scan was used for identification the specific product ions from each of the precursors. Suggested structures consistent with the product ions produced are depicted for each analyte.

Fig. 2. Precursor ion →product ion transitions and suggested structures of the major product ion derived from linoleic acid and its oxidation products.

Fig. 2

Open in a new tab

ESI negative-ion full scan was used for identification of precursor ions for each of the analytes and ESI negative-ion product scan was used for identification the specific product ions for each of the precursors. Suggested structures consistent with the product ions produced are depicted for each analyte.

MATERIALS

The following compounds were purchased from Cayman Chemical Company (Ann Arbor, MI):

  • (±)-9-hydroxy-10E,12Z-octadecadienoic acid (9-HODE) catalog number 38400

  • (±)-13-hydroxy-9Z,11E-octadecadienoic acid (13-HODE) catalog number 38600

  • 9-oxo-10E,12Z-octadecadienoic acid (9-oxoODE) catalog number 38420

  • 13-oxo-9Z,11E-octadecadienoic acid (13-oxoODE) catalog number 38620

  • (±)5-hydroxy-6E,8Z,11Z,14Z-eicosatetraenoic acid (5-HETE) catalog number 34210

  • (±)8-hydroxy-5Z,9E,11Z,14Z-eicosatetraenoic acid (8-HETE) catalog number 34340

  • (±)-9-hydroxy-5Z,7E,11Z,14Z-eicosatetraenoic acid (9-HETE) catalog number 34400

  • (±)11-hydroxy-5Z,8Z,12E,14Z-eicosatetraenoic acid (11-HETE) catalog number 34500

  • (±)12-hydroxy-5Z,8Z,10E,14Z-eicosatetraenoic acid (12-HETE) catalog number 34550

  • (±)15-hydroxy-5Z,8Z,11Z,13E-eicosatetraenoic acid (15-HETE) catalog number 34700

  • (±)5(6)-epoxy-8Z,11Z,14Z-eicosatrienoic acid (5,6-EET) catalog number 50211

  • (±)8(9)-epoxy-5Z,11Z,14Z-eicosatrienoic acid (8,9-EET) catalog number 50351

  • (±)11(12)-epoxy-5Z,8Z,14Z-eicosatrienoic acid (11,12-EET) catalog number 50511

  • (±)14(15)-epoxy-5Z,8Z,11Z-eicosatrienoic acid (14,15-EET) catalog number 50651

  • 5-oxo-6E,8Z,11Z,14Z-eicosatetraenoic acid (5-oxoETE) catalog number 34250

  • 12-oxo-5Z,8Z,10E,14Z-eicosatetraenoic acid (12-oxoETE) catalog number 34580

  • 15-oxo-5Z,8Z,11Z,13E-eicosatetraenoic acid (15-oxoETE) catalog number 34730

  • 9α,11α,15S-trihydroxy-5Z,13E-dien-1-oic acid (PGF2α) catalog number 16010

  • 9Z,12Z-octadecadienoic acid (linoleic acid, LA) catalog number 90150

  • 5Z,8Z,11Z,14Z-eicosatetraenoic acid (arachidonic acid, AA) catalog number 90010

  • 15S-hydroxy-5Z,8Z,11Z,13E-eicosatetraenoic-5,6,8,9,11,12,14,15-d8 acid (15S-HETE-d8) catalog number 334570

  • 9α,11α,15S-trihydroxy-5Z,13E-dien-1-oic-3,3,4,4-d4 acid (PGF2α-d4) catalog number 316350

  • 2,3-dinor-8-iso Prostaglandin F2α(2,3-dinor-8-isoPGF2α)catalog number16290

  • 8-iso Prostaglandin F2α (8-isoPGF2α) catalog number16350

  • 5-isoprostane F2α (5-iPF2α-VI) catalog number16300

  • Prostaglandin F2α (PGF2α) catalog number16010

The following chemicals and solvents were purchased from Fisher Scientific (Pittsburg, PA):

  • Methanol (Optima grade) catalog number A454

  • Water (HPLC grade) catalog number W5

  • Acetic Acid (glacial, ACS grade) catalog number A38

  • 2-Propanol (HPLC grade) catalog number A451

  • Hexane (HPLC grade) catalog number H302

  • Sodium Hydroxide (NaOH, ACS grade) catalog number S318

  • Hydrochloric acid (HCl, ACS grade) catalog numberA144

The following reagents were purchased from Sigma Aldrich (St. Louis, MO):

  • Tin (II) chloride dihydrate (SnCl2, 98%, ACS grade) catalog number 243523

  • 2,6-Di-tert-butyl-4-methylphenol (BHT, >99%) catalog number B1378

  • Argon (5.0 Ultra High Purity, part number AR5.0UH) was obtained from Praxair Inc. (Cleveland, OH)

  • Nitrogen (5.0 Ultra High Purity, part number NI5.0UH) was obtained from Praxair Inc. (Cleveland, OH)

INSTRUMENTATION

An AB SCIEX API 4000 triple quadrupole mass spectrometer interfaced to a Shimadzu HPLC system through an electrospray ionization source controlled by a PC is used for single column developmental work. Alternatively, for high-throughput dual high pressure binary gradient elution (Fig. 8) we used a Shimadzu Prominence HPLC system (Shimadzu Scientific Instruments, Columbia, MD) composed of four Shimadzu LC-20AD pumps, a Shimadzu DGU-20A5 vacuum degasser and a SIL-HTC autosampler with three, 2-position, 6 port, high pressure switching valves attached. For higher sensitivity we use an AB SCIEX (Framingham, MA) API 5000 mass spectrometer in the negative ESI mode. We required a centrifuge capable of spinning sample tubes at a relative centrifugal force of 2000 × g, and a vortex mixer (Fisher Scientific MaxiMix catalog number 12-815-50 or equivalent). Organic solutions are measured using a syringe (Gastight, Hamilton Company series 1700) or a bottle top dispenser (Fisher Scientific, catalog number 03-692-178). Plasma, urine and other aqueous solutions are measured using positive displacement pipettors (Fisher Scientific catalog number 21-377-328).

Fig. 8. Dual High Pressure Mixed Binary Gradient Elution HPLC System.

Fig. 8

Open in a new tab

The plumbing schematic for the HPLC system used for high-throughput chromatography is shown. The valves are configured so that column 2 is in-line with the mass spectrometer while valve 3 serves as a bypass valve to waste, preventing effluent from reaching the mass spectrometer during any period of non-data collection.

PROTOCOL

Standard solutions and plasma and urine sample preparation

Stock solutions of all the standards were prepared in 85% methanol. Glass containers were flushed with argon and kept at −20°C. Our data indicate that the standards solutions are chemically stable for at least six months under these conditions. Working solutions were prepared monthly by diluting the stock solutions into 85% methanol. To prepare the standard curves, 10 μl of internal standard solution containing 20 ng/μl PGF2α-d4 and 4 ng/μl 15(S)-HETE-d8 was added into 200 μl of each dilution of the standard solutions. The entire process is detailed below:

  1. Aliquot 200μl of plasma (or urine) sample into a clean 12 × 75mm borosilicate glass test tube

  2. Add 10μl of internal standard mixture (20 ng/μl PGF2α-d4 and 4 ng/μl 15(S)-HETE-d8)

  3. Add 1.0ml 10% v/v acetic acid in water/2-propanol/hexane (2/20/30, v/v/v)

  4. Vortex briefly to mix

  5. Add 2.0mL Hexane

  6. Cap tube with polypropylene stopper (Fisher Scientific catalog #14-376-77)

  7. Vortex mix for three (3) minutes

  8. Centrifuge sample at room temperature 2000 × g for five (5) minutes

  9. Remove upper hexane layer with a glass Pasteur pipette; place in clean 12 × 75mm test tube

  10. Evaporate hexane extract under N2 flow.
    1. For unesterified fatty acid analysis: proceed with step 26.

  11. For total fatty acids profile; suspend pellet in 1.0ml of 2-propanol

  12. Add 1.0ml of aqueous 2M NaOH (saturated with Argon)

  13. Vortex briefly to mix

  14. Blanket sample with Argon gas

  15. Cap tube immediately with polypropylene stopper (Fisher Scientific catalog #14-376-77)

  16. Place in 60°C water bath for 60min to hydrolyze lipids

  17. Cool to room temperature
    1. At this point samples may be stored refrigerated overnight if necessary

  18. Acidify sample by adding 1.2 ml of 2.0 M HCl in water

  19. Add 2.0mL Hexane

  20. Cap tube with polypropylene stopper (Fisher Scientific catalog #14-376-77)

  21. Vortex mix for three (3) minutes

  22. Centrifuge sample at room temperature 2000 × g for five (5) minutes

  23. Remove upper hexane layer with a glass Pasteur pipette; place in clean 12 × 75mm test tube

  24. Repeat steps 19 through 23 and combine both hexane layers

  25. Evaporate hexane extract under N2 flow

  26. Vortex pellet into 100 μl of 85% v/v methanol in water

  27. Store under Argon at −20°C until analysis by LC-MS/MS

  28. Centrifuge at 2000 × g immediately before analysis
    1. For oxidized fatty acids and their precursor compounds in urine, the same protocol is used as described for plasma above except that a more sensitive mass spectrometer is needed for the analysis.

LC-MS/MS analysis

A 40μl sample (in 85% methanol) was injected onto a reverse-phase C18 HPLC column (XPERTEX, 2.1 × 250mm, 5 μm particle, P.J. Cobert Associates, St. Louise, Missouri) at flow rate of 0.2 ml/min. Solvent A was water (plus 0.2% v/v acetic acid) and solvent B was methanol (plus 0.2% v/v acetic acid). The column was equilibrated with 85% B. The separation was performed starting at 85% B for 10 min, then using a gradient to 100% B over two min, followed by 100% B for 10 min and then 100% A for 10 min. HPLC column effluent was introduced into an AB SCIEX (Framingham, MA) API 4000 triple quadrupole mass spectrometer. The source of the mass spectrometer was configured with the electrospray needle voltage set at −4.2 kV, and a Turbo ion spray temperature of 350 °C. Nitrogen gas was used for the nebulizer, curtain and collision gas. The gas flow rates, collision energies, declustering potentials, entrance potentials, focusing potentials, and collision cell exit potentials were adjusted to give optimal signal response from the direct infusion of a dilute solution of each standard in aqueous methanol. Analyses were performed using electrospray ionization in negative-ion mode with MRM of precursor and characteristic product ions specific for each analyte monitored for 100 milliseconds per transition. The mass-to-charge ratio (m/z) transitions monitored are shown in Figures 1 and 2. The MRM transitions for two internal standards was m/z 327 → 182 for 15(S)-HETE-d8 and m/z 357 → 197 for PGF2α-d4. The internal standard PGF2α-d4 was used for quantification of PGF2α and 15(S)-HETE-d8 was used for quantification of EETs, HETEs, HODEs, oxoETEs, oxoODEs, linoleic acid, and arachidonic acid. In cases where greater sensitivity was needed for detection, such as in measuring urine F2isoprostanes, an AB SCIEX API 5000 triple quadrupole mass spectrometer was used.

CALCULATIONS and EXPECTED RESULTS

A typical LC-MS/MS chromatogram for analysis of 15 different fatty acid derivatives including two internal standards using MRM mode is shown in Fig. 3. Although the retention time and the precursor ions for 9-HODE and 13-HODE are nearly the same, they can still be quantified using specific product ions (m/z 171 for 9-HODE and m/z 195 for 13-HODE).

Fig. 3. Negative ion LC-MS/MS chromatography of selected un-oxidized fatty acids and oxidized fatty acid standards.

Fig. 3

Open in a new tab

Chromatographic separation was performed on a reverse phase C18 column (2.1 × 250mm, 5μm particle) using acidified methanol/water as the mobile phase at flow rate of 0.2 ml/min using a gradient as described in the protocol.

Calibration curves and limit of detection (LOD)

Calibration curves for all the fatty acid derivatives monitored, relative to their selected internal standard, are shown in Fig. 4. Serial dilutions of authentic fatty acid derivatives including arachidonic acid, linoleic acid and their oxidation products (HETEs, EETs, oxoETEs, HODEs and oxoODEs) in an 85% methanol solution demonstrate a linear response over a wide range of the standard concentrations throughout the assay range (Figs. 4A–4C). Analyte recovery was quite high with more than 95% for arachidonic acid and linoleic acid, and more than 85% for their oxidation products except for PGF2α which had only 10% recovery. Use of the stable isotope synthetic internal standard for PGF2α allows us to account for the reduced lipid extraction characteristics of this oxidized lipid species. Fig. 5 shows the chromatography and calibration curve of authentic PGF2α relative to its internal standard PGF2α-d4. Because biomarkers of oxidation accumulate only to trace amounts in biological samples, sensitivity is critical. We therefore determined the limit of detection (LOD) and limit of quantitation (LOQ) of LC-MS/MS for the species being monitored. The LOD was expressed as the lowest amount of analyte on column generating a signal-to-noise ratio of at least three. The LOQ was examined by reducing the concentration of standard solution gradually and is expressed as the lowest concentration yielding a signal-to-noise ratio of at least 10. As shown in Table 1, the most readily detected analyte is 11-HETE, with LOD<2.6pg and LOQ<0.09 ng/ml. The LOD and LOQ provided are instrument specific. The instrument selected for these analyses is relatively old, and not of the highest sensitivity. Even with the abundance of the molecular species being monitored in plasma or urine this protocol allows for ready detection and quantification even with almost decade old triple quadrupole mass spectrometers.

Fig. 4. Internal standard calibration curves for the un-oxidized fatty acids and oxidized fatty acids analyzed using LC-MS/MS.

Fig. 4

Open in a new tab

15(s)-HETE-d8 (12 fmol on column) was used as internal standard for preparing these standard calibration curves. Calibration range: 0 – 4 mol/mol, Panel A) 15(S)-HETE-d8 for HETEs, Panel B) EETs, oxoEETs, Panel C) HODEs, and oxoODEs, Panel D) 18 – 150 mol/mol 15(S)-HETE-d8 for linoleic acid (LA) and arachidonic acid (AA).

Fig. 5. Chromatography and internal standard calibration curve for PGF2α analyzed using LC-MS/MS.

Fig. 5

Open in a new tab

PGF2α-d4 (60 fmol on column) was used as internal standard for preparing the standard calibration curve for PGF2α. Panel A) precursor to product ion chromatogram, Panel B) calibration curve, calibration range: 0 – 4 mol/mol PGF2α-d4.

Table 1.

Limit of Detection (LOD) and Limit of Quantitation (LOQ)

Analyte Transition (m/z) aLOD (pg) bLOQ (ng/ml)
Arachidonic acid 303 → 259 91 3.1
5-HETE 319 → 115 4.8 0.16
8-HETE 319 → 155 4.8 0.1
9-HETE 319 → 151 9.6 0.30
11-HETE 319 → 167 2.6 0.09
12-HETE 319 → 179 3.2 0.11
15-HETE 319 → 175 13 0.45
5,6-EET 319 → 191 16 0.54
8,9-EET 319 → 151 19 0.64
11,12-EET 319 → 167 16 0.54
14,15-EET 319 → 175 26 0.90
5-oxoETE 317 → 203 32 1.1
12-oxoETE 317 → 153 32 1.1
15-oxoETE 317 → 113 64 2.2
PGF2α 353 → 193 7.1 0.24
Linoleic acid 279 → 261 617 21
9-HODE 295 → 171 15 0.50
13-HODE 295 → 195 30 1.0
9-oxoODE 293 → 185 47 1.6
13-oxoODE 293 → 113 29 1.0
Open in a new tab
a

LOD was defined as a peak whose signal-to-noise ratio is at 3:1.

b

LOQ was defined as a peak whose signal-to-noise ratio is at 10:1

Precision and Accuracy

Precision and accuracy of the intraday assay were assessed using standards dissolved in PBS buffer that were assayed in replicates (n=4) for known concentrations of analytes. All samples were analyzed on the same day and their back-calculated concentrations were determined from the internal calibration curves prepared the same day. The interday assay precision and accuracy were assessed by assaying known concentrations of analytes in replicates (n=3) over three different days. The precision was expressed as the coefficient of variation (%) and the accuracy as the percentage bias (%). For the intraday and interday assay (Table 2) the CV% and % bias for almost all of the analytes are substantially less than 10%, and typically in the 3–5% range.

Table 2.

Intraday and Interday Assay Precision and Accuracy

Analytes Concentration (ng/ml)
Intraday Assay (n=4) Interday Assay (n=9)
Added Mean SD CV (%) Bias (%) Mean SD CV (%) Bias (%)
Arachidonic acid 300 313 10 3.3 4.4 280 9.5 3.4 −6.8
5-HETE 15.0 5.5 1.4 9.3 2.8 15.0 1.2 7.7 −0.1
8-HETE 15.0 16.0 1.3 8.3 6.4 15.6 1.2 7.8 4.0
9-HETE 15.0 15.5 2.1 14 3.4 15.0 1.6 10.9 0.1
11-HETE 15.0 15.0 0.6 4.3 0.2 14.7 0.8 5.4 −2.2
12-HETE 15.0 15.4 0.5 3.4 2.7 14.9 0.8 5.3 −3.9
15-HETE 15.0 15.2 1.2 7.7 1.2 15.0 0.8 5.6 −0.1
PGF2α 5.0 5.5 0.6 10 9.6 5.4 0.6 10.6 7.8
Linoleic acid 450 473 16 3.3 5.0 431 23 5.3 −4.2
9-HODE 15.0 15.2 1.1 7.2 1.3 14.9 0.8 5.5 −0.5
13-HODE 15.0 15.3 0.6 3.8 1.7 15.1 0.6 4.0 0.8
Open in a new tab

LC-MS/MS analysis of oxidized fatty acids in plasma

This LC-MS/MS method can be used to identify and quantify fatty acid derivatives in biological matrices like plasma, serum or other tissues. Fig. 6 shows an example of quantification of multiple oxidized and un-oxidized fatty acid derivatives in 0.2 ml human plasma. This method clearly identified all the peaks of analytes monitored in plasma. The levels of some oxidized and un-oxidized fatty acids in human plasma are shown in Table 3.

Fig. 6. Representative LC-MS/MS chromatogram of un-oxidized fatty acids and oxidized fatty acids extracted from human plasma.

Fig. 6

Open in a new tab

Extraction as described in “Materials and Methods”; separation was performed on a reverse phase C18 column (2.1 × 250mm, 5μm) using acidified methanol/water as mobile phase at flow rate of 0.2 ml/min using a gradient as described in “Materials and Methods”. Note that some of the oxidized fatty acid species have common MRM pairs and unique retention times;8,9-EET exhibits the same MRM transitions as both 8-HETE and 9-HETE, and 11,12-EET shares the same MRM transitions with both 11-HETE and 12-HETE. The unique retention times of the 8,9-EET and 11,12-EET were established by injection of authentic standards (data not shown).

Table 3.

Levels of Selected Oxidized Fatty Acids and Their Precursors in Human Plasma from Healthy Volunteers

Fatty acids ng/ml plasma
Mean (n=15) SD
Arachidonic acid 638 228
5-HETE 2.9 1.4
8-HETE 2.5 1.1
9-HETE 3.3 1.3
11-HETE 3.2 1.1
12-HETE 12 3.5
15-HETE 4.8 2.0
PGF2α 4.1 1.0
Linoleic acid 2124 946
9-HODE 16 5.0
13-HODE 47 20
Open in a new tab

LC-MS/MS analysis of F2-isoprostanes in urine

Determination of isoprostanes in urine provides a superior non-invasive methodology to assay the index of oxidative stress[2, 8, 9, 12, 14, 19, 23]. Since urine contains markedly less arachidonic acid (and linoleic acid) and other interfering compounds than plasma or serum, these samples are much less prone to the artificial generation of interfering oxidized fatty acids.

CAVEATS

Sample Collection

For sample collection it is important to minimize any extraneous oxidation that may occur after its initial collection. Storage conditions must be carefully chosen to exclude air which can be a primary source of contaminating oxidant. It is important to blanket biological tissues and fluids (like serum or plasma) with an inert atmosphere such as nitrogen or, preferably, with heavier than air argon gas to displace the oxygen around the sample. Ideally, antioxidants are included in the liquid portion of the sample. Commonly used antioxidants are free radical scavenging compounds like fat soluble butylated hydroxytoluene (BHT) and divalent metal ion scavengers. The choice of divalent metal ion scavengers is critical to avoid extraneous oxidation; pentacoordinate chelators like diethylenetriaminepentaacetic acid (DTPA), which envelope divalent ions and do not leave open coordination sites available where catalysis can occur, are the reagents of choice for this purpose. Tetra-coordinated divalent metal ion species in complexes with chelators like ethylenediaminetetraacetic acid (EDTA) and ethyleneglycoltetraacetic acid (EGTA) have open sites that can promote oxidation, and even paradoxically enhance the redox potential of the active site on the divalent metal ion[38–40]. In addition, sample storage at low temperatures is crucial. Many laboratories use ultralow (−70°C to −80°C) freezers for this purpose. Experience has shown that lipids in biological samples survive long term storage better at these lower temperatures. It is also important to limit the number of freeze thaw cycles a sample will be exposed to during its lifetime. Smaller volume, one-time use, aliquots are preferred for storage, since oxidized lipid recovery is degraded in biological samples through multiple freeze thaw cycles. Indeed, it is wise to aliquot a biological specimen into multiple vials destined for lipid analysis, overlay the samples with argon, and then “snap freeze” these portions in liquid nitrogen immediately after collection and processing.

Fatty Acid Extraction

Antioxidants used in sample preservation

To quantify total free oxidized and un-oxidized fatty acids, lipids in the sample are hydrolyzed to their free fatty acid components with base at elevated temperature before extraction. Antioxidants need to be included in the hydrolysis medium; these can be BHT and DTPA as described above. Sometimes it is helpful to also include a mild reducing agent like tin chloride (SnCl2), dithionate or borohydrides (sodium borohydride and sodium cyanoborohydride for example). If using a mild reducing agent, we prefer tin chloride since all of the others have partial solubility in isopropyl alcohol and may carry through the extraction procedure, and both interfere with ionization efficiency and, over time, contaminate the mass spectrometer source (i.e. tin chloride does not partition much into the hexane layer of the Hara extraction). With reducing reagents, labile species like lipid hydroperoxides and cyclic endoperoxides are reduced to their corresponding hydroxyl compounds[32]. Obviously, any assay that requires separate determination of the hydroperoxide or endoperoxide oxidation products must avoid the use of reducing agents like tin chloride. It should be noted that tin chloride reduction is the classic way to produce prostaglandin D2 from PGH2, raising the possibility of artificial elevation in F2-Isoprostane levels in plasma with addition of tin chloride[41]. However, in our hands, control studies have shown that addition of tin chloride to plasma does not increase measured F2-Isoprostane levels, presumably because the level of hydroperoxide present in plasma is exceedingly low relative to the F2-Isoprostane level. To analyze isoprostanes in plasma it has been shown by others that one should first add reducing agent (to eliminate unstable hydroperoxides and cyclic endoperoxides) and then hydrolyze for optimal results [42]. In our opinion, analyses of urine F2-isoprostane is the superior method for global oxidative stress measures since there is very little precursor arachidonate or hydroperoxide/endoperoxide present, and artifactual formation of the oxidized lipid during prolonged storage does not occur. In plasma matrix, we have found improved recovery of some un-oxidized fatty acids (arachidonic and linoleic acid) with shorter hydrolysis times (30 minutes) at higher temperatures (60°C), whereas hydrolysis for PGF2α is optimal at longer times (60 minutes) at lower temperatures (45°C). The choice of base used for saponification and its concentration are also important. Generally sodium hydroxide is used, which is easily neutralized with acid to form a water soluble salt. Optimization of the conditions of hydrolysis for the analyte of interest needs to be done using authentic standards.

As mentioned above, antioxidants may need to be incorporated into one or both phases of the extraction solvents. Years of experience have shown us that it is inefficient to include a derivatization step for any compound unless the derivatives are absolutely necessary to provide the sensitivity for the detection of the analytes. Most modern mass spectrometers have the sensitivity to measure compounds down to the femtomole and even attomole range. Careful choice of sample amounts and their dilutions coupled with the resolution and concentrative properties of high performance liquid chromatography should allow for easy detection of most oxidized fatty acid derivatives and their parent compounds.

Fatty Acid Analysis

Which fatty acids to monitor?

The last step is the separation and quantitation of the oxidized fatty acid derivatives and their parent compounds from the sample matrix. Which fatty acid and oxidized fatty acid species to monitor is largely dependent on the sample matrix and scientific question? While cycloxoygenase (COX) derived oxidized species like PGF2α and lipoxygenase products like HETEs and HODEs predominate in plasma, non-enzymatic free radical-derived oxidized species like the F2-isoprostanes predominate in urine. As far as biological mechanisms and outcomes are concerned, there may be times when it is desirable to monitor both groups of products in either matrix[3, 6]. As a global measure of oxidative stress in a patient it is most desirable to use a non-invasive assay such as urinalysis of oxidized lipid species. The invasive plasma (or serum) matrix determination of oxidized lipid species is useful for the evaluation of COX pathway activity, as in inflammatory disease states like arthritis and atherosclerosis. In urine the stable end products of the free radical oxidized fatty acids, such as 2,3-dinor-8-isoprostaglandin F2α, predominate.

Generally speaking, the oxidation products from arachidonic acid can be found in human urine whereas the linoleate derivatives are typically present at insignificant amounts, often approximating analytical noise levels. The two-step free radical oxidation products from arachidonic acid, the isoprostanes and PGF2α, are the most abundant and most persistent (stable). The 2,3–dinor-8-isoprostaglandin F2α, a metaobolite of 8-isoprostaglandin F2α, and 5-isoprostglandin F2α predominate in urine from humans (see Fig. 7). In fact, stability studies with prolonged incubations of urine at 37° C reveal that addition of an antioxidant cocktail as a sample preservative may not even be necessary. Urine specimens have proven to be stable at room temperature with respect to almost all oxidized fatty acid species examined. The stability has been observed for a number of hours, and even many days in specific instances; as expected[2], inclusion of the antioxidant mixtures mentioned above markedly increases this stability. F2-isoprostanes at room temperature are stable for days. What is important is to obtain a clean catch of urine that is free of cell debris or bacteria. Most samples analyzed from males are of this nature; however those from females can sometimes contain cellular material rich in the precursor arachidonic acid. This interference can mostly be eliminated from all samples by briefly spinning them in a centrifuge before aliquoting for storage. Visible hematuria and urinary tract infections may present samples that yield abnormally high levels of oxidation products. Control studies have found that spot urine collection provides levels of F2-isoprostanes on par with those obtained from a 24-hour collection. The dilution levels of these oxidized species from urine are corrected with respect to the urine creatinine level and are expressed as mass of oxidized or un-oxidized fatty acid to mass of creatinine. A clean catch spot urine analysis of oxidized fatty acids like F2-isoprostanes can provide an immediate non-invasive global systemic index of oxidative stress.

Fig. 7. A representative LC-MS/MS chromatogram of PGF2α's extracted from human urine.

Fig. 7

Open in a new tab

Top panel shows heavy isotope labeled internal standard PGF2α-d4 from urine second panel shows the corresponding transition for PGF2α's which are chromatographically resolved from their free radical derived isomers, the 8-Iso-PGF2α's. Bottom two panels show other free radical derived PGF2α isomers with different MRM transitions.

Standards

The standards themselves are carefully chosen, based on their chemical structure, to avoid deuterium exchange with the solvents. When possible, the concentrations of the stable isotope labeled internal standards are verified against primary standards, typically from the linearity plots of their unlabeled isotopologues. Care must be taken to include an amount of stable isotope labeled unsaturated parent fatty acid (arachidonic or linoleic acid) in the general concentration range of endogenous fatty acid derivative levels. If oxidation is observed in this internal standard channel for a particular sample; it could be re-prepared and re-run. At worst, if all samples show some degree of artificial oxidation, the extraction method itself may need to be modified by incorporating either better or higher concentrations of antioxidants. Use of distinct stable isotope labeled precursor to monitor for potential artificial oxidation (by monitoring of potential MRM of heavy isotope labeled isotopologues) is critical for assay development to ensure the assay is not artificially generating the oxidation products monitored in the endogenous sample. Absolute quantitation of the contribution from artifactual oxidation products to samples should be less than 10% (maximum) of the amount measured.

QC Standards for Fatty Acid Analysis Performance

It is helpful to first run a solution containing both the analytes and internal standards in the same solvent mixture used for the sample dissolution but not in the sample matrix (i.e. in a simple solvent system). This standard is used primarily to assess the suitability of the chromatographic system to perform the assay. If peaks are poorly resolved, out of sequence, or absent, the run can be stopped immediately and the chromatographic issues addressed without losing a precious sample. The sample matrix itself presents a formidable challenge to the chromatography system as well as the mass spectrometer. It is wise to prepare a significant volume of control standard consisting of analytes spiked into the matrix of interest at physiologically relevant concentrations (typically at the midpoint of the compound's standard curve). Several of these control standards are processed in parallel with the samples (at least one for every five to 10 samples). Minimally, this check standard is analyzed first and a blank is run immediately thereafter. These high-throughput analyses demand a robust HPLC system with an autosampler allowing for the storage of a run-full of samples (up to 100 per day) under refrigeration.

HPLC System Hardware

There are a plethora of HPLC systems available for this task from numerous manufacturers. Herein, we explore just a few of the options and alternatives. Nano-flow HPLC (flow less than 1 μl/min)offers the best chromatographic effects and sensitivity, but suffers from its expense, fragility and ease of contamination. One stray dust particle is enough to clog a flow path and shut the system down for a major prolonged and expensive repair. It is our opinion that it is best to reserve nano-flow for stringent analyses requiring the utmost in sensitivity, not routine high-throughput analyses. Next is micro and or narrow bore chromatography (flow rates from 1 μl/min to 200 μl/min). This is the optimal operational region for a high-throughput chromatography system even though 2.1 mm diameter columns are starting to be considered old technology. Most HPLC hardware developed over the last 20 years is able to operate reliably in this flow window, and most importantly, most mass spectrometers have heated sources that can handle these flow rates without splitting. At 200 μl/min flow rate, solvents are still used slowly, with amounts of less than 300 ml of solvent consumption and waste per system per day. A note on Ultrahigh Pressure Liquid Chromatography (UPLC): while smaller particle sizes in stationary phases offer enhanced rapid resolution of components, the chromatographic resolution can make peaks of interest so thin that they cannot be properly digitized by the mass spectrometry system. This can be especially true for cases like these analyses where many different MRM channels of data must be simultaneously collected during the chromatographic run. Large scale studies and the desire for robust reproducible performance with few difficulties (e.g. 300 to 600 sample injections before needing to replace the column) make the high pressure mixing pump configuration at a 200 μl/min flow rate our preferred system since it provides adequate sensitivity and resolution for the analytes being quantified. Normal bore chromatography with flow rates greater than 200 μl/min and up to 1 ml/min is now old technology with solvent usage and waste amounts of up to a liter and a half per day. Chromatographic dilution at these higher flow rates and inefficient vaporization and sampling of the resulting spray in the mass spectrometer source attenuates signals, making these high flow methods less desirable.

Multiplexing Considerations

Efficient use of mass spectrometer time may be further optimized in this multiplexing approach when three HPLC systems and columns are used. The actual advantage of a three column versus two column system amounts to an approximate 10% time savings. This only becomes important when thousands of samples must be analyzed under the same conditions on the same data acquisition system. There are multiplexed HPLCs consisting of four (or more) chromatographic systems run in parallel in a “Gatling gun” fashion to maintain 100% useable time on the mass spectrometer. While such systems sound ideal, it is our experience that a four-fold multiplex system is not typically necessary. The system complexity, cost and maintenance become increasingly prohibitive as each chromatographic system is added. Typically, chromatography conditions can be fine-tuned to allow for faster overall runs, and a two- or three-column multiplex system can provide near optimal instrument efficiency. Additionally, we have found that column manufacturers often do not supply separation columns that demonstrate the same chromatographic performance that column switching requires. Even consecutive columns from the same lot will not provide the same performance (same peak retention times, peak resolution and tailing factor). Consequently, we purchase one dozen or more of the same column type at a time and screen all the columns with standards to identify pairs of columns that work best with each other in terms of retention characteristics (times and peak shapes). This process greatly facilitates the programming of the multiplex HPLC system. When some columns in each purchase appear to have no matching partner, these unusable columns can usually be returned for credit, provided only a single clean standard solution has been injected, and the column has been well maintained.

Another issue with multiplexed HPLC systems is high pressure versus low pressure mixing with the HPLC pumps. Low pressure mixing, where the solvents are mixed by a proportioning valve located ahead (upstream) of the pumps, is much simpler to maintain and operate since only one pump per analytical column is required to deliver the gradient. However, low pressure pumps have larger delay volumes with a resulting longer gradient delay. They also cannot generate as reproducible a gradient as a high pressure mixing system. High pressure mixing requires two pumps, one for the aqueous phase and one for the organic phase, with their flow outputs connected together on the high pressure side of the pumps. Differential flow rates through the two pumps generate the gradient and since pump flows can be tightly controlled, the gradients supplied by binary high pressure systems are very accurate.

We have settled on a dual high pressure binary gradient elution system for all multiplexed high-throughput applications, like oxidized fatty acid analyses[23, 27, 28, 33]. Maintaining four HPLC pumps and one autosampler with two or three switching valves is a tenable goal (Fig. 8 illustrates the plumbing for a 2-HPLC multiplex system). The heart of the dual column switching system are two 6-port two position valves that simultaneously switch one HPLC system or the other in-line with the autosampler and mass spectrometer (Fig. 8). The dual column advantage is that data can immediately be collected from a second column after gradient elution and concurrent data collection from the first column. The chromatographic run on the second column takes place while the first column is being washed, re-equilibrated and re-injected. In a single column system the mass spectrometer would not be collecting any useful data during the column washout re-equilibration and sample injection steps. Even in the most rapid chromatographic gradient separations, the dual column approach will offer this time savings in the dwell time of column washout, re-equilibration and subsequent sample injection. One has to decide if the added complexity is worth the increase in speed. In our experience, column switching increases sample throughput since it takes significantly more time to wash, re-equilibrate, re-inject and wait for the salt front to elute from a reverse phase chromatography column compared to the time it takes to run the analytical part of the gradient where peaks of interest elute and data is collected. Specifically, in the analysis of the oxidized and un-oxidized fatty acids, the mass spectrometer idle time (during the 10 minutes of column equilibration back into the aqueous phase and the two minutes of the sample injection and salt front washout) are utilized for data collection from the second column. Since the actual data collection occurs over 22 minutes (of a total run time of 34 minutes) this translates into an increase in used mass spectrometer time of 12 minutes per sample. When thousands of clinical samples must be analyzed 12 minutes equates to 23 more samples that can be run per day or a difference between 24 days (no multiplexing) and 15 days (with multiplexing) per 1000 samples.

More complicated front end chromatography systems with more than two binary pump HPLC systems require a more complicated system of valves including a multi-position selector valve. We have chosen to incorporate a bypass valve into our chromatographic systems that directs the HPLC stream to waste during any time where there is no data collection occurring[28], which helps keep the mass spectrometer and it's source clean. This setup also requires that the autosampler or system controller have suitable relay type outputs for controlling the valves and starting data collection on the mass spectrometer. The dual high pressure binary systems we use have four contact closure outputs used for this purpose: one to start data collection on the mass spectrometer, two to control the valves that switch between the HPLC systems and columns, and one to control the additional bypass valve. Valves such as these that can accommodate lower flow rates (in the nano-flow range) and higher pressures (UPLC) are also more expensive to purchase and maintain.

Mass Spectrometer Considerations

We purchase mass spectrometers based on their sensitivity to the compounds in our assay. It is necessary to be able to detect a number of MRM pairs to accurately quantify the compounds of interest. With an average peak width of 30 seconds in our chromatographic runs, a data sampling rate of 1.5 seconds (the time it takes for the mass spectrometer to cycle through every MRM pair) provides adequate digitization (more than 15 data points over each peak). For a typical Gaussian shaped peak, we recommend collecting at least 10 to 12 data points over the duration of the peak for accurate digitization[43]. More sensitive high end instruments allow for faster sampling rates permitting the use of thinner (low volume) chromatographic peaks, like those from UPLC. The trade-off for the thinner UPLC peaks is a faster, more sensitive mass spectrometry instrument with the possibility of some uncertainty in the quantitation if an adequate data sampling rate can barely be maintained. Faster scan rates at higher resolutions on newer instrumentation increases the amount of MRM data that can be collected over time. Higher order scanning (MS3, MS4, etc.) aids in structural confirmation, though sometimes at the expense of time needed to collect enough data for accurate MRM quantitation. Widely available triple quadrupole instruments, due to their specificity and sensitivity are a must. We envision each MRM pair as a specific detector for each individual compound. Thus, if we require confirmation of the identity of a compound we typically collect more than one MRM transition for that compound. Matching chromatographic retention time and the ratio of the signals between two (or more) distinct MRM channels for the same analyte is a powerful argument to confirm identity. This also brings to light the issues with the sensitivity range that can be addressed by the careful selection of MRM pairs. Some fatty acids, like the precursor (un-oxidized) species (e.g. linoleic and arachidonic acid), are present at much higher concentrations than their oxidized products. Selection of an alternate MRM pair where the precursor or product ion is not the base peak (at 100% intensity) allows for the quantitation of compounds when one or another MRM pair reaches saturation levels in the mass spectrometer detector. This lower abundance channel is then used for quantitation. Linearity is determined as a ratio of the concentration to a stable isotope labeled fatty acid of similar structure, and the peak area ratio of the MRM pair from the native fatty acid derivative to the corresponding MRM pair from the isotopologue standard.

SUMMARY/CONCLUSION

Rapid and efficient high-throughput LC-MS/MS analysis of oxidized fatty acid derivatives is feasible. The sample preparation entails addition of stable isotope synthetic standards, basic hydrolysis and liquid/liquid phase extraction. Specific oxidized free fatty acids and their precursors in the extract are separated through a reverse phase C18 column using acidified methanol/water as the mobile phase and detected by multiple reaction monitoring using negative ion ESI/MS/MS. In the examples provided, more than 20 different oxidized fatty acids and their precursors were analyzed, including hydroxy-octadecadienoic acids (HODEs), oxo-octadecadienoic acids (oxoODEs), hydroxy-eicosatetraenoic acids (HETEs), epoxy-eicosatrienoic acids (EETs), oxo-eicosatetraenoic acids (oxo-ETEs), 9α, 11α, 15S-trihydroxy-5Z, 13E-dien-1-oic acid (PGF2α), and their precursors (linoleic acid and arachidonic acid). Internal standard PGF2α-d4 was used for the calibration of PGF2α and 15-HETE-d8 was used for the calibration of the other lipids monitored. This LC-MS/MS method enables high sample throughput (more than 20 specific lipid peroxidation products may be analyzed from a single injection) and is optimal for gaining a complex lipid oxidation profile. The recovery of oxidized fatty acid derivatives extracted from plasma under the methods employed is typically over 85% except for PGF2α. All the aforementioned specific lipid peroxidation products have been detected in human plasma from healthy subjects and found to be above the limit of quantitation. For intraday precision, CVs ranged from 3.3% to 10% and accuracy ranged from 99 to 109%. For the interday precision, CVs ranged from 3.4 to 11% and accuracy ranged from 93 to 108%. This LC-MS/MS method is specific, sensitive, reproducible, and practical for analysis of lipid peroxidation products in human plasma, and F2-isoprostane analyses in urine.

Ox FA Method Highlights

  • Multiplex LC/MS/MS analyses are efficient for oxidized fatty acid quantification in clinical samples

  • Non–invasive measurement of F2Isoprostanes in urine is as a general measure of oxidative stress

  • High throughput liquid chromatography tandem mass spectrometry assay for oxidized fatty acids

  • Immediate non-invasive global systemic index of patient oxidative stress

ACKNOWLEDGMENTS

Mass Spectrometry instrumentation used was housed within the Cleveland Clinic Mass Spectrometry Facility with partial support through a Center of Innovation by AB SCIEX. This work was supported by National Institutes of Health grants P01HL076491, P01HL098055, P20HL113452, P01HL103453, R01HL103866, S10RR21029, S10RR025548 and S10RR026738. ZW is partially supported by an AHA Scientist Development Grant SDG12050473. SLH is also supported by a gift from the Leonard Krieger Fund.

Abbreviations

AA

arachidonic acid

BHT

butylated hydroxytoluene

DTPA

diethylenetriaminepentaacetic acid

ETEs eicosatrienoic acids

EETs, epoxy-eicosatrienoic acids

ESI

electrospray ionization

ESI−

electrospray negative ionization

PGF2α

9α,11α,15S-trihydroxy-5Z,13E-d8acid-1-oic acid

15(S)-HETE-d8

15S-hydroxy-5Z,8Z,11Z,13E-eicosatetraenoic-5,6,8,9,11,12,14,15-d8acid

H(P)ETE

hydroxy-eicosatetraenoic acid and hydroperoxy-eicosatetraenoic acid

HPLC

high performance liquid chromatography

H(P)ODE

hydroxy-octadecadienoic acid and hydroperoxy-octadecadienoic acid

LC-MS/MS

liquid chromatography with electrospray ionization on-line tandem mass spectrometry

LA

linoleic acid

LOD

limit of detection

MS

mass spectrometry

MRM

multiple-reaction monitoring

oxoETEs

oxo-eicosatetraenoic acids

oxoODEs

oxo-octadecadienoic acids

PGF2α-d4

9α,11α,15S-trihydroxy-5Z,13E-dien-1-oic-3,3,4,4-d4acid

See materials section for more specific abbreviations of oxidized fatty acid species.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • [1].Shishehbor MH, Hazen SL. Antioxidant studies need a change of direction. Cleve Clin J Med. 2004;71:285–288. doi: 10.3949/ccjm.71.4.285. [DOI] [PubMed] [Google Scholar]
  • [2].Bastani NE, Gundersen TE, Blomhoff R. Dried blood spot (DBS) sample collection for determination of the oxidative stress biomarker 8-epi-PGF(2alpha) in humans using liquid chromatography/tandem mass spectrometry. Rapid Commun Mass Spectrom. 2012;26:645–652. doi: 10.1002/rcm.6149. [DOI] [PubMed] [Google Scholar]
  • [3].Brose SA, Thuen BT, Golovko MY. LC/MS/MS method for analysis of E(2) series prostaglandins and isoprostanes. J Lipid Res. 2011;52:850–859. doi: 10.1194/jlr.D013441. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [4].Dickinson JS, Murphy RC. Mass spectrometric analysis of leukotriene A4 and other chemically reactive metabolites of arachidonic acid. J Am Soc Mass Spectrom. 2002;13:1227–1234. doi: 10.1016/S1044-0305(02)00456-7. [DOI] [PubMed] [Google Scholar]
  • [5].Dzeletovic S, Breuer O, Lund E, Diczfalusy U. Determination of cholesterol oxidation products in human plasma by isotope dilution-mass spectrometry. Anal Biochem. 1995;225:73–80. doi: 10.1006/abio.1995.1110. [DOI] [PubMed] [Google Scholar]
  • [6].Feldstein AE, Lopez R, Tamimi TA, Yerian L, Chung YM, Berk M, Zhang R, McIntyre TM, Hazen SL. Mass spectrometric profiling of oxidized lipid products in human nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. J Lipid Res. 2010;51:3046–3054. doi: 10.1194/jlr.M007096. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [7].Hazen SL, Zhang R, Shen Z, Wu W, Podrez EA, MacPherson JC, Schmitt D, Mitra SN, Mukhopadhyay C, Chen Y, Cohen PA, Hoff HF, Abu-Soud HM. Formation of nitric oxide-derived oxidants by myeloperoxidase in monocytes: pathways for monocyte-mediated protein nitration and lipid peroxidation In vivo. Circ Res. 1999;85:950–958. doi: 10.1161/01.res.85.10.950. [DOI] [PubMed] [Google Scholar]
  • [8].Hellmuth C, Weber M, Koletzko B, Peissner W. Nonesterified fatty acid determination for functional lipidomics: comprehensive ultrahigh performance liquid chromatography-tandem mass spectrometry quantitation, qualification, and parameter prediction. Anal Chem. 2012;84:1483–1490. doi: 10.1021/ac202602u. [DOI] [PubMed] [Google Scholar]
  • [9].Janicka M, Kubica P, Kot-Wasik A, Kot J, Namiesnik J. Sensitive determination of isoprostanes in exhaled breath condensate samples with use of liquid chromatography-tandem mass spectrometry. J Chromatogr B AnalytTechnol Biomed Life Sci. 2012;893–894:144–149. doi: 10.1016/j.jchromb.2012.03.005. [DOI] [PubMed] [Google Scholar]
  • [10].Jiang X, Ory DS, Han X. Characterization of oxysterols by electrospray ionization tandem mass spectrometry after one-step derivatization with dimethylglycine. Rapid Commun Mass Spectrom. 2007;21:141–152. doi: 10.1002/rcm.2820. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [11].Meagher EA, Barry OP, Lawson JA, Rokach J, FitzGerald GA. Effects of vitamin E on lipid peroxidation in healthy persons. JAMA. 2001;285:1178–1182. doi: 10.1001/jama.285.9.1178. [DOI] [PubMed] [Google Scholar]
  • [12].Medina S, Dominguez-Perles R, Gil JI, Ferreres F, Garcia-Viguera C, Martinez-Sanz JM, Gil-Izquierdo A. A ultra-pressure liquid chromatography/triple quadrupole tandem mass spectrometry method for the analysis of 13 eicosanoids in human urine and quantitative 24 hour values in healthy volunteers in a controlled constant diet. Rapid Commun Mass Spectrom. 2012;26:1249–1257. doi: 10.1002/rcm.6224. [DOI] [PubMed] [Google Scholar]
  • [13].Podmore ID, Griffiths HR, Herbert KE, Mistry N, Mistry P, Lunec J. Vitamin C exhibits pro-oxidant properties. Nature. 1998;392:559. doi: 10.1038/33308. [DOI] [PubMed] [Google Scholar]
  • [14].Sala A, Kayganich K, Zirrolli JA, Murphy RC. Negative ion tandem mass spectrometry of leukotriene E4 and LTE4 metabolites: identification of LTE4 in human urine. J Am Soc Mass Spectrom. 1991;2:314–321. doi: 10.1016/1044-0305(91)80023-Z. [DOI] [PubMed] [Google Scholar]
  • [15].Schmitt D, Shen Z, Zhang R, Colles SM, Wu W, Salomon RG, Chen Y, Chisolm GM, Hazen SL. Leukocytes utilize myeloperoxidase-generated nitrating intermediates as physiological catalysts for the generation of biologically active oxidized lipids and sterols in serum. Biochemistry. 1999;38:16904–16915. doi: 10.1021/bi991623w. [DOI] [PubMed] [Google Scholar]
  • [16].Shishehbor MH, Aviles RJ, Brennan ML, Fu X, Goormastic M, Pearce GL, Gokce N, Keaney JF, Jr., Penn MS, Sprecher DL, Vita JA, Hazen SL. Association of nitrotyrosine levels with cardiovascular disease and modulation by statin therapy. JAMA. 2003;289:1675–1680. doi: 10.1001/jama.289.13.1675. [DOI] [PubMed] [Google Scholar]
  • [17].Shishehbor MH, Brennan ML, Aviles RJ, Fu X, Penn MS, Sprecher DL, Hazen SL. Statins promote potent systemic antioxidant effects through specific inflammatory pathways. Circulation. 2003;108:426–431. doi: 10.1161/01.CIR.0000080895.05158.8B. [DOI] [PubMed] [Google Scholar]
  • [18].Steffenrud S, Borgeat P, Evans MJ, Bertrand MJ. Mass spectrometry of prostaglandins, leukotrienes and steroids as their allyldimethylsilyl ether derivatives. Biomed Environ Mass Spectrom. 1986;13:657–661. doi: 10.1002/bms.1200131204. [DOI] [PubMed] [Google Scholar]
  • [19].Sterz K, Scherer G, Ecker J. A simple and robust UPLC-SRM/MS method to quantify urinary eicosanoids. J Lipid Res. 2012;53:1026–1036. doi: 10.1194/jlr.D023739. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [20].Strife RJ, Simms JR. Mass spectrometry/mass spectrometry of prostaglandins: Daughter ion spectra of derivatized and isotope-labeled E and D prostanoids. Anal Chem. 1988;60:1800–1807. doi: 10.1021/ac00168a030. [DOI] [PubMed] [Google Scholar]
  • [21].Thomas SR, Stocker R. Molecular action of vitamin E in lipoprotein oxidation: implications for atherosclerosis. Free RadicBiol Med. 2000;28:1795–1805. doi: 10.1016/s0891-5849(00)00236-7. [DOI] [PubMed] [Google Scholar]
  • [22].Tsikas D. Application of gas chromatography-mass spectrometry and gas chromatography-tandem mass spectrometry to assess in vivo synthesis of prostaglandins, thromboxane, leukotrienes, isoprostanes and related compounds in humans. J Chromatogr B Biomed SciAppl. 1998;717:201–245. doi: 10.1016/s0378-4347(98)00210-2. [DOI] [PubMed] [Google Scholar]
  • [23].Zhang H, Il'yasova D, Sztaray J, Young SP, Wang F, Millington DS. Quantification of the oxidative damage biomarker 2,3-dinor-8-isoprostaglandin-F(2alpha) in human urine using liquid chromatography-tandem mass spectrometry. Anal Biochem. 2010;399:302–304. doi: 10.1016/j.ab.2009.12.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [24].Zhang R, Brennan ML, Shen Z, MacPherson JC, Schmitt D, Molenda CE, Hazen SL. Myeloperoxidase functions as a major enzymatic catalyst for initiation of lipid peroxidation at sites of inflammation. J BiolChem. 2002;277:46116–46122. doi: 10.1074/jbc.M209124200. [DOI] [PubMed] [Google Scholar]
  • [25].Zhang R, Shen Z, Nauseef WM, Hazen SL. Defects in leukocyte-mediated initiation of lipid peroxidation in plasma as studied in myeloperoxidase-deficient subjects: systematic identification of multiple endogenous diffusible substrates for myeloperoxidase in plasma. Blood. 2002;99:1802–1810. [PubMed] [Google Scholar]
  • [26].Davies SS, Pontsler AV, Marathe GK, Harrison KA, Murphy RC, Hinshaw JC, Prestwich GD, Hilaire AS, Prescott SM, Zimmerman GA, McIntyre TM. Oxidized alkyl phospholipids are specific, high affinity peroxisome proliferator-activated receptor gamma ligands and agonists. J BiolChem. 2001;276:16015–16023. doi: 10.1074/jbc.M100878200. [DOI] [PubMed] [Google Scholar]
  • [27].Hishinuma T, Suzuki K, Saito M, Yamaguchi H, Suzuki N, Tomioka Y, Kaneko I, Ono M, Goto J. Simultaneous quantification of seven prostanoids using liquid chromatography/tandem mass spectrometry: the effects of arachidonic acid on prostanoid production in mouse bone marrow-derived mast cells. Prostaglandins LeukotEssent Fatty Acids. 2007;76:321–329. doi: 10.1016/j.plefa.2007.04.005. [DOI] [PubMed] [Google Scholar]
  • [28].Kita Y, Takahashi T, Uozumi N, Shimizu T. A multiplex quantitation method for eicosanoids and platelet-activating factor using column-switching reversed-phase liquid chromatography-tandem mass spectrometry. Anal Biochem. 2005;342:134–143. doi: 10.1016/j.ab.2005.03.048. [DOI] [PubMed] [Google Scholar]
  • [29].Podrez EA, Poliakov E, Shen Z, Zhang R, Deng Y, Sun M, Finton PJ, Shan L, Febbraio M, Hajjar DP, Silverstein RL, Hoff HF, Salomon RG, Hazen SL. A novel family of atherogenic oxidized phospholipids promotes macrophage foam cell formation via the scavenger receptor CD36 and is enriched in atherosclerotic lesions. J BiolChem. 2002;277:38517–38523. doi: 10.1074/jbc.M205924200. [DOI] [PubMed] [Google Scholar]
  • [30].Podrez EA, Poliakov E, Shen Z, Zhang R, Deng Y, Sun M, Finton PJ, Shan L, Gugiu B, Fox PL, Hoff HF, Salomon RG, Hazen SL. Identification of a novel family of oxidized phospholipids that serve as ligands for the macrophage scavenger receptor CD36. J BiolChem. 2002;277:38503–38516. doi: 10.1074/jbc.M203318200. [DOI] [PubMed] [Google Scholar]
  • [31].Roberts JL, Morrow JD. Analysis of F2-Isoprostanes by Gas Chromatography-Mass Spectrometry/Negative Ion Chemical Ionization. Methods in Biological Oxidative Stress. 2003;1:33–39. [Google Scholar]
  • [32].Guichardant M, Chen P, Liu M, Calzada C, Colas R, Vericel E, Lagarde M. Functional lipidomics of oxidized products from polyunsaturated fatty acids. ChemPhys Lipids. 2011;164:544–548. doi: 10.1016/j.chemphyslip.2011.05.002. [DOI] [PubMed] [Google Scholar]
  • [33].Oertel R, Richter K, Fauler J, Kirch W. Increasing sample throughput in pharmacological studies by using dual-column liquid chromatography with tandem mass spectrometry. J Chromatogr A. 2002;948:187–192. doi: 10.1016/s0021-9673(01)01458-3. [DOI] [PubMed] [Google Scholar]
  • [34].Folch J, Lees M, Sloane Stanley GH. A simple method for the isolation and purification of total lipides from animal tissues. J BiolChem. 1957;226:497–509. [PubMed] [Google Scholar]
  • [35].Bligh EG, Dyer WJ. A rapid method of total lipid extraction and purification. Can J BiochemPhysiol. 1959;37:911–917. doi: 10.1139/o59-099. [DOI] [PubMed] [Google Scholar]
  • [36].Hara A, Radin NS. Lipid extraction of tissues with a low-toxicity solvent. Anal Biochem. 1978;90:420–426. doi: 10.1016/0003-2697(78)90046-5. [DOI] [PubMed] [Google Scholar]
  • [37].Ottesen LH, Harry D, Frost M, Davies S, Khan K, Halliwell B, Moore K. Increased formation of S-nitrothiols and nitrotyrosine in cirrhotic rats during endotoxemia. Free RadicBiol Med. 2001;31:790–798. doi: 10.1016/s0891-5849(01)00647-5. [DOI] [PubMed] [Google Scholar]
  • [38].Balcerczyk A, Sowa K, Bartosz G. Metal chelators react also with reactive oxygen and nitrogen species. Biochemical and Biophysical Research Communications. 2007;352:522–525. doi: 10.1016/j.bbrc.2006.11.053. [DOI] [PubMed] [Google Scholar]
  • [39].Engelmann MD, Bobier RT, Hiatt T, Cheng IF. Variability of the Fenton reaction characteristics of the EDTA, DTPA, and citrate complexes of iron. BioMetals. 2003;16:519–527. doi: 10.1023/a:1023480617038. [DOI] [PubMed] [Google Scholar]
  • [40].Li L, Abe Y, Kanagawa K, Shoji T, Mashino T, Mochizuki M, Tanaka M, Miyata N. Iron-chelating agents never suppress Fenton reaction but participate in quenching spin-trapped radicals. AnalyticaChimicaActa. 2007;599:315–319. doi: 10.1016/j.aca.2007.08.008. [DOI] [PubMed] [Google Scholar]
  • [41].Hamberg M, Samuelsson B. Detection and Isolation of an Endoperoxide Intermediate in Prostaglandin Biosynthesis. ProcNatlAcadSci. 1973;70:899–903. doi: 10.1073/pnas.70.3.899. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [42].Milne GL, Musiek ES, Morrow JD. F2-Isoprostanes as markers of oxidative stress in vivo: An overview. Biomarkers. 2005;10:10–23. doi: 10.1080/13547500500216546. [DOI] [PubMed] [Google Scholar]
  • [43].Rainville PD. Addressing the Issues of Very Sharp LC Peaks Within Quantitative LC/MS/MS. Chromatography Today. 2010;3:12–14. [Google Scholar]

RESOURCES