|Home | About | Journals | Submit | Contact Us | Français|
Prostanoids are potent mediators of many physiological and pathophysiological processes. Of the many analytical methodologies used for their qualitative and quantitative analysis, electrospray tandem mass spectrometry coupled to liquid chromatography (ESI-LC-MS/MS) offers a rapid, sensitive and versatile system applicable to lipidomic analyses. We have developed an ESI-LC-MS/MS assay for twenty-seven mediators including prostaglandins, prostacyclines, thromboxanes, dihydroprostaglandins and isoprostanes. The assay was liner over the concentration range 1-100 pg/μL. The limits of detection and quantitation were 0.5-50 pg and 2-100 pg respectively, whilst recoveries were from 83-116 % depending on the metabolite. The assay can be applied to profiling prostanoids produced by a variety of biological fluids and extracts including brain, liver, plasma and urine, facilitating thus our understanding of the role of these lipid mediators in health and disease, as well as assisting in drug development.
Prostanoids, a term that collectively describes prostaglandins, prostacyclins and thromboxanes, are a sub-class of the lipid mediator group known as eicosanoids . They derive from C-20 polyunsaturated fatty acids, mainly dihomo-γ-linoleic (20:3n-6), arachidonic (20:4n-6), and eicosapentaenoic (20:5n-3) acids, through the action of cyclooxygenases-1 and -2 (COX-1 and COX-2)  (Figure 1 and scheme 1). The reaction product of COX is the unstable endoperoxide prostaglandin H (PGH) that is further transformed to the individual prostanoids by a series of specific prostanoid synthases . Prostanoids are local-acting mediators formed and inactivated within the same or neighbouring cells prior to their release into circulation as inactive metabolites (15-keto- and 13,14-dihydro-keto metabolites) . Non-enzymatic peroxidation of arachidonic acid and other fatty acids in vivo can result in prostaglandin-like substances isomeric to the COX-derived prostaglandins that are termed isoprostanes (scheme 1c) .
Prostanoids take part in many physiological and pathophysiological processes in practically every organ, tissue and cell, including the vascular, renal, gastrointestinal and reproductive systems [1,6-9]. Their activities are mediated through prostanoid-specific receptors and intracellular signalling pathways, whilst their biosynthesis and action are blocked by nonsteroidal anti-inflammatory drugs (NSAID). Isoprostanes are considered to be reliable markers of oxidant stress status and have been linked to inflammation, ischaemia-reperfusion, diabetes, cardiovascular disease, reproductive disorders, diabetes .
Methods currently used for the analysis of prostanoids include HPLC with fluorescence or UV detection, GC, LC-MS, GC-MS, enzyme immunoassays and radioimmunoassays [1,11]. Although the immunoassays are very popular, they have low specificity and are not applicable to the simultaneous profiling of more than one metabolites at a time. HPLC and GC-MS methods offer greater sensitivity and flexibility but require derivatisation or the use of radiolabelled fatty acid precursors to facilitate detection and increased sensitivity, since most prostanoids do not absorb UV except at low wavelengths . When analysed by electrospray ionisation (ESI), eicosanoids easily form protonated and deprotonated molecules . Consequently, the use of tandem mass spectrometry (MS/MS) coupled to liquid chromatography (LC-MS/MS) can lead to the development of fast and sensitive analytical protocols with high-throughput potential. Furthermore, LC-MS/MS assays permit the simultaneous profiling and quantitative analysis of prostanoids, dihydroprostaglandins and isoprostanes without the need of derivatisation and lengthy sample preparation, thus overcoming the limitations of GC-MS, HPLC, and other conventional methodologies.
Lipidomics is an emerging area of lipid research aiming to study the profiles and complete composition of lipids in parallel with their functional role in any given tissue or system [14,15]. The study of prostanoid profiles in conjunction with other metabolic profiles, enzyme activities, protein and gene expression can offer valuable insights to tissue and organ function in health and disease. In this study we present a rapid method for the simultaneous qualitative and quantitative assay of twenty-seven prostanoids, dihydroprostanoids and isoprostanes. This method can be applied to elucidate the profiles of prostanoids and their inactivated metabolites in different cellular systems and body fluids, and consequently assist any lipidomic or systems biology application aiming to explore physiological states, disease phenotypes, novel biomarkers or the development of therapeutic approaches.
Prostaglandin D1 (PGD1), prostaglandin E1(PGE1), prostaglandin F1α (PGF1α), 6-keto-prostaglandin F1α (6-keto-PGF1α), prostaglandin B2 (PGB2), prostaglandin B2-d4 (PGB2-d4), prostaglandin D2 (PGD2), prostaglandin E2 (PGE2), prostaglandin F2α (PGF2α), prostaglandin J2 (PGJ ), Δ12 2-Prostaglandin J (Δ12 2-PGJ ), 15-deoxy-Δ12,142-prostaglandin J (15-deoxy-Δ12,14 2-PGJ2), prostaglandin D3 (PGD3), prostaglandin E3 (PGE3), prostaglandin F3α (PGF3α), thromboxane B2 (TXB2), thromboxane B3 (TXB3), 8-iso-PGE2, 8-iso-PGF2α, 8-iso-15-keto PGE2α, 8-iso-15-keto PGF2α, 13,14-dihydro PGE1, 13,14-dihydro PGF1α, 13,14-dihydro PGF2α, 13,14-dihydro-15-keto PGE1, 13,14-dihydro-15-keto PGF1α, 13,14-dihydro-15-keto PGE2 and 13,14-dihydro-15-keto PGF2α were purchased from Cayman Chemicals (Ann Arbor, MI, USA). HPLC grade solvents, glacial acetic acid and all other chemicals were from Sigma-Aldrich (Dorset, UK). Solid phase extraction (SPE) cartridges (C18-E 500 mg, 6 mL) were purchased from Phenomenex (Macclesfield, UK).
Stock standard solutions of all prostanoids and isoprostanes were prepared in ethanol (400 pg/μL) and stored at -20°C under N2. Composite standard solutions were prepared by mixing and diluting the appropriate stock solutions to the final concentrations of 100 pg/μL, 50 pg/μL, 20 pg/μL, 10 pg/μL and 1 pg/μL. The internal standard (PGB2-d4) was prepared in ethanol (1 ng/μL) and added to all composite standards at a final concentration of 400 pg/μL. The peak-area ratios of every compound to PGB2-d4 were calculated and plotted against the concentration of the calibration standards. Calibration lines were calculated by the least squares linear regression method. To calculate the concentration of any given analyte the peak-area ratio to PGB2-d4 was calculated and read off the corresponding calibration line. The limit of detection was calculated by using a signal to noise ratio of 3. The limit of quantitation was determined by using a signal to noise ratio of 10. Peak integrations and signal to noise calculations were performed using the MassLynx™ V4.0 software (Waters) using the manufacturers’ instructions.
Brain, liver, plasma and urine samples were collected from male Wistar rats. Tissue samples (approximately 500 mg) were homogenised in water (35 up and down strokes) using a Dounce glass mini homogeniser (2 mL) with tight fitting pestle. During this process the homogeniser was kept on ice. The resulting solution was adjusted to 15% methanol (v/v) (final volume 3 mL). Plasma and urine samples (500 μL) were diluted with water and adjusted to 15 % methanol (v/v), to a final volume of 3 mL. Internal standard PGB2-d4 (40 ng) was added to each sample. The samples were incubated on ice for 30 min and then centrifuged at 3000 rpm for 5 min to remove any precipitated proteins. The resulting clear supernatants were acidified with 0.1 M hydrochloric acid to pH 3.0 and immediately applied to SPE cartridges that had been preconditioned with 20 mL methanol followed with 20 mL water. The cartridges were then washed with 20 mL 15% (v/v) methanol, 20 mL water, and 10 mL hexane in succession. Finally, the prostanoids were eluted with 15 mL methyl formate. The extraction procedure was performed using a vacuum manifold (Phenomenex); the vaccum was adjusted so that individual drops could be seen from each cartridge. The organic solvent was evaporated under a fine stream of nitrogen, the residue was dissolved in 100 μL ethanol and stored at -20°C awaiting LC-MS/MS analysis.
The LC-MS/MS analysis was performed on a Waters Alliance 2695 HPLC pump coupled to an electrospray (ESI) triple quadrupole Quattro Ultima mass spectrometer (Waters, Elstree, Hertsfordshire, UK). Instrument control and data acquisition were performed using the MassLynx™ V4.0 software. The instrument was operated in the negative ionisation mode. For optimisation of MS and MS/MS conditions, standards at a concentration of 10 ng/μL were individually introduced to the spectrometer by direct infusion through a syringe pump (flow rate of 10 μL/min) into the HPLC solvent flow (flow rate 0.2 mL/min). The capillary voltage was set at 3500 V, source temperature 120 °C, desolvation temperature 360 °C, and cone voltage 35 V. The collision energy was optimised for each compound to get optimum sensitivity using argon as collision gas. Dwell times were 0.2 s. The collision energy settings used for each one of the 21 MRM transitions are summarised in Table 1.
Chromatographic analysis was performed on a C18 column (Gemini, 5μ, 150 x 2 mm) (Phenomenex, Macclesfield, UK). Sample injections were performed with a Waters 2690 autosampler and the sample chamber temperature was set at 8 °C. The injection volume was 5 μL and the flow rate 0.2 mL/min. The column was maintained at ambient temperature. The analysis was performed using an acetonitrile-based gradient system mixing two solvents: Solvent A was acetonitrile : water : glacial acetic acid, 45:55:0.02 (v/v/v); Solvent B was acetonitrile : water : glacial acetic acid, 90:10:0.02 (v/v/v). The analytes were separated using the following gradient: 0.0-8.0 min, 0% solvent B; 8.0-8.1 min, 0 to 50 % solvent B; 8.0-12.0 min 50% solvent B; 12.0-12.1 min, 50 to 70 % solvent B; 12.1-20.0 min 70 % solvent B; 20.0-20.1 min, 70 to 0 % solvent B; 20.1-30.0 min 0% solvent B.
Rat brain, liver, plasma and urine samples were spiked with 200 pg of a mixed prostanoid and isoprostane standard. The metabolites were extracted as described above, and analysed in parallel with the extracts of un-spiked tissue samples to estimate the recovery. The recovery of 8-iso-PGE2 was calculated at the absence of PGE2. In detail: the extracts of un-spiked tissue samples allowed us to estimate the peak-area value corresponding to the amount of naturally occurring metabolites in those tissues. This figure was then subtracted from the corresponding peak-area value obtained from the spiked samples. The difference is equivalent to the amount of metabolite added to each tissue sample. This figure was then compared to the peak-area value that was obtained from the analysis of meabolites that did not undergo extraction thus representing 100% of the initial concentration. The resulting value shows the recovery of each metabolite expressed as percent of the initial concentration. Three separate sets of recovery experiments were performed per type of tissue or body fluid.
Prostanoid profiles are tissue and stimuli-dependent and cannot be predicted. Furthermore, quantitative methodologies require versatility, high sensitivity and selectivity because of the low concentrations, short half-lives and structural similarities of these metabolites . Electrospray ionisation (ESI) has been readily applied in the analysis of all classes of eicosanoids. Since prostanoids and isoprostanes have free carboxylic acid moieties, ES-results in an abundant [M-H]- carboxylate ion that allows for relatively low concentrations to be detected . Multiple reaction monitoring assays (MRM) allow for the further improvement of the detection and quantitation limits of LC-MS assays. So far, only a small number of prostanoids, dihydroprostaglandins and isoprostanes have been studied together for the development of LC-MS/MS assays that detect more than one class of eicosanoids [13,17-20]. We have developed a sensitive ESI-LC-MS/MS assay applicable to the simultaneous detection and quantitation of twenty-seven eicosanoids including sixteen prostanoids (scheme 1a and b), seven dihydroprostanoids and four isoprostanes (scheme 1c).
In order to develop the LC-MS/MS assay, production scanning experiments were conducted using argon as collision gas and the collision energy was optimised for each individual compound to generate the most abundant product ions. These product ion spectra were then used to select the precursor-product ion pairs for the development of MRM assays (spectra presented as Supplementary Data, see attached file). Table 1 summarises the optimal collision energies and ion pairs selected for the ESI-LC-MS/MS assay. The choice of product ions for MRM was in agreement with the available literature values, i.e.: PGE2 251→271, PGF2α 353→193; 6-keto-PGF1α 369→163, PGE3 349→269, TXB2 369→169; 8-iso-PGF2α 353→193; 8-iso-15-keto PGF2α 351→315 [17-20]. Although PGD3 has been detected using the transition 349→233 , we opted for the transition 349→269 that is the same one used to detect PGE3. The isobaric compounds PGD3 and PGE3 are chromatographically separated therefore this choice of transition allows the use of only one MRM to detect both compounds (Figure 2A). Furthermore, the product ion m/z 269 is of higher abundance in the relevant spectra under the conditions used in our study. Overall, the analysis of all the twenty-seven metabolites included in the present study was performed using twenty-one MRM transitions (Table 1).
The compounds were chromatographically resolved on a C18 column using a gradient of two acetonitrile-based solvents. Figure 2 shows representative chromatograms of the analysis of these lipid mediators, i.e. 16 prostanoids including prostaglandins, cyclopentanone prostaglandins and thromboxanes, 7 dihydroprostanoids and 4 isoprostanes. The run time of the assay was 30 min including a 10 min wash cycle programmed to run before the next injection. Since the retention time (r.t) of the last eluting compound was 18 min (Figure 2A), the wash cycle and therefore the run time of the assay can be further reduced depending on the application (e.g. quality of biological sample; need for high-throughput). Overall, this run time is comparable to other HPLC-based reported methods of eicosanoid analysis and provides the basis for a rapid assay [16,21].
The co-eluting prostanoids PGE1 and PGE2 (r.t. 4.47 min, Figure 2A and 2B) were detected separately using two different MRM transitions (i.e. 353→317 and 351→271 respectively). Conversely, the isobaric compounds PGE1 and PGD1 are identified through the same MRM transition but were separated chromatographically (353→317; r.t. 4.47 and 4.79 min respectively; Figure 2B). The same principle was applied to PGE2 and PGD2 (351→271; r.t.: 4.47 and 5.18 min respectively), PGE3 and PGD3 (349→269; r.t. 3.85 and 4.08 min respectively) and PGJ2 and Δ12-PGJ2 (333→271; r.t. 9.17 and 9.80 min respectively) (Figure 2A), and PGF2α and its isomer 8-iso-PGF2α (353→193; r.t. 3.38 and 3.77 respectively, Figure 2B). The tailing peak observed for 6-keto-PGF1α (Figure 2B) is a result of the chromatography and not the MS assay (the single standard gives a chromatographic peak of similar shape and it is not contaminated with impurities). Although this badly shaped peak may interfere with the reproducibility of the integration, it has not affected the linearity of the calibration line (Table 2).
The two isobaric isoprostanes 8-iso-15-keto PGF2α and 8-iso-PGE2 presented an extra chalenge: Although the transition 351→315 is the most selective for both compounds, the product ion m/z 315 is also generated from PGE2 and PGD2. As a result this transition brings up all four compounds when simultaneously present in the sample (Figure 2A). Whereas 8-iso-15-keto PGF2α, PGE2 and PGD2 are chromatographically well resolved, 8-iso-PGE2 coelutes with PGE2. Therefore, the identification of the isoprostane 8-iso-PGE2 may not be possible at the presence of PGE2. Further development of the gradient system could resolve the issue of separating those two isomers and thus overcome this limitation of the assay.
Standard calibration lines were typically constructed for each analyte over the range of 5 to 500 pg per injection. Calibration lines were calculated by the least squares linear regression method and the results confirmed that the assay was linear over this range (Table 2). In order to assess the recovery of the extraction methodology, rat brain, liver, plasma and urine samples were spiked with mixed standards (200 pg/compound) and analysed. The peak area for each extracted analyte was compared with the one for non-extracted standard from which recovery was calculated. The average recoveries ranged from 83 to 116 % depending on the compound and type of biological sample (Results summarised in Table 3). Finally, the limit of detection (LoD) was found to be in the range of 0.5-50 pg whilst the limit of quantitation (LoQ) was in the range of 2-100 pg, depending on the compound (results are summarised in Table 2). Overall, the assay presented in this paper has been found to perform well, showing good linearity, recovery and sensitivity.
The reported ESI-LC-MS/MS assay can be used to identify and quantify a large number of prostanoids in different types of biological samples. The method is simple, sensitive, high-throughput and lends itself to lipidomic applications.
This work has been partially supported by The Wellcome Trust, Grant WT077714. The authors thank Andrew Healey (Analytical Centre, University of Bradford) for excellent technical assistance and Dr N Blagden for useful advice.