An improved LC-MS/MS procedure for brain prostanoid analysis using brain fixation with head-focused microwave irradiation and liquid-liquid extraction.

High-performance liquid chromatography with tandem mass spectrometry detection (LC-MS/MS) allows a highly selective, sensitive, simultaneous analysis for prostanoids (PG) without derivatization. However, high chemical background noise reduces LC-MS/MS selectivity and sensitivity for brain PG analysis. Four common methods using different solvent systems for PG extraction were tested. Although these methods had the same recovery of PG, the modified acetone extraction followed by liquid/liquid purification had the greatest sensitivity. This method combined with hexane/2-propanol extraction permits the simultaneous analysis of other lipid molecules and PG in the same extract. We also determined that PG mass in brain powder stored at -80 degrees C was reduced 2- to 4- fold in 4 weeks; however, PG were stable for long periods (>3 months) in hexane/2-propanol extracts. PG mass was increased significantly when mice were euthanized by decapitation and the brains rapidly flash-frozen rather than euthanized using head-focused microwave irradiation. This reduction is not the result of PG trapping or destruction in microwave-irradiated brains, demonstrating its importance in limiting mass artifacts during brain PG analysis. Our improved procedure for brain PG analysis provides a reliable, rapid means to detect changes in brain PG mass under both basal and pathological conditions and demonstrates the importance of sample preparation in this process.

plication of LC-MS/MS requires special care for brain PG extraction and sample preparation before analysis, because of the complicated brain biological matrix that produces high chemical background noise, thereby reducing the selectivity and sensitivity of detection. Several different methods are used for brain PG extraction, including methanol extraction followed by solid-phase extraction (9,12,13), hexane/2-propanol extraction (14), ether extraction (15), and acetone/chloroform extraction (16). Although methanol extraction followed by solid-phase extraction (9) and ether extraction (15) have been used in different LC-MS/MS procedures, it is important to note that the simultaneous evaluation of these different extraction methods for use in brain PG analysis using LC-MS/MS has not been done.
Another important factor that must be considered during brain PG analysis is the method used to euthanize the animal and the subsequent handling of the brain sample. Brain PG mass in rodents euthanized by headfocused microwave irradiation is 10-to 40-fold lower than in animals euthanized by decapitation (17,18). Although it is assumed that this reduction in PG mass is the result of heat inactivation of the enzymes involved in postmortem PG formation (17)(18)(19), this reduction could also be the result of the trapping or destruction of PG in microwaved brain. Several lines of evidence support the assumption that PG are not trapped or destroyed during microwave treatment. Brain PG mass found in indomethacin-treated animals euthanized by decapitation does not differ from PG mass in brains from animals euthanized by microwave irradiation (18). Also, after intracerebral ventricular injection of radiolabeled PG before microwave irradiation, most of the recovered radioactivity was in the form of PG; however, the recovery of the radiolabeled PG from brains subjected to microwave irradiation was not examined in this study (19). Importantly, these studies do not provide direct evidence that PG are not trapped or destroyed in microwaved brains.
In the present study, we evaluated and modified existing methods for brain PG extraction and sample preparation for LC-MS/MS analysis. The modified method improved the limits of tissue PG detection by 4-to 20-fold in an individual PG-dependent manner and allowed the analysis of PG in ,10 mg of brain tissue with an extraction recovery that ranged from 85% to 95%. We also evaluated PG stability during storage and analysis and provide direct evidence that PG are not trapped or destroyed in microwaved brains.

Animals
This study was conducted in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals (Publication 80-23) and under an animal protocol approved by the Institutional Animal Care and Use Committee at the University of North Dakota (Protocol 0409-9). Male 129/SvEv strain mice (25-30 g) were maintained on standard laboratory chow diet and water ad libitum. The ages of the mice used in this study were between 9 and 11 months.

Brain PG extraction
Fasted male mice were anesthetized with halothane (1-3%) and euthanized using head-focused microwave irradiation (2.8 kW, 1.35 s; Cober Electronics, Inc., Norwalk, CT) to heatdenature enzymes in situ. The whole brain was removed, frozen in liquid nitrogen, and pulverized under liquid nitrogen temperatures to a fine, homogeneous powder.
The extraction protocol was a modification of a previously published procedure (16) that was adapted for tissue extraction. Pulverized tissue (10-20 mg) was homogenized in 3 ml of acetone-saline (2:1) containing PGE 2 -d4 and 6-oxo-PGF 1a -d4 (100 pg in 10 ml of acetonitrile) as internal standards and 0.005% butylated hydroxytoluene (BHT) to prevent PG oxidation using a Tenbroeck tissue grinder (Kontes Glass Co., Vineland, NJ). The homogenate was transferred to a screw-top tube, vortexed for 4 min, and subjected to 10 min of centrifugation (2,000 g) at 4jC. The supernatant was transferred to another screw-top tube and mixed with 2.0 ml of hexane by vortexing for 0.5 min. Then, the mixture was subjected to 10 min of centrifugation (2,000 g) at 4jC. The upper phase containing hexane with extracted lipids was discarded, the lower phase was acidified with formic acid to pH 3.5 (30 ml of 2 M formic acid), and 2 ml of chloroform containing 0.005% BHT was added. The mixture was vortexed for 0.5 min and again subjected to 10 min of centrifugation (2,000 g) at 4jC to aid in the separation of the two phases. The lower phase containing chloroform was transferred to a screw-top tube silanized with Sigmacote: (Sigma Chemical Co., St. Louis, MO), flushed with nitrogen, and cooled at 220jC for at least 2 h. This cooling allows the separation of any residual upper phase, which is then removed and discarded before analysis.

Sample preparation for LC-MS/MS
After the residual upper phase was discarded, 200 ml of methanol was added to the extract and it was dried down under a stream of nitrogen. The dried extract was transferred to 100 ml silanized microvial inserts (National Scientific, Rockwoods, TN; catalog No. C4010-S630) using 2 3 0.1 ml of chloroform containing 10% methanol and 0.005% BHT. The solvent in microvial inserts was dried down under a stream of nitrogen. The transfer procedure was repeated twice. Ten microliters of acetonitrile was added to the insert with dried extract, vortexed for 30 s, and mixed with 20 ml of water.

Reverse-phase HPLC electrospray ionization mass spectrometry
The separation was carried out using a Luna C-18(2) column (3 mm, 100 Å pore diameter, 150 3 2.0 mm; Phenomenex, Torrance, CA) with a stainless-steel frit filter (0.5 mm) and security guard cartridge system (C-18) (Phenomenex). The HPLC system consisted of an Agilent 1100 series LC pump equipped with a wellplate autosampler (Agilent Technologies, Santa Clara, CA). The autosampler was set at 4jC. Twenty-five microliters out of a 30 ml sample was injected onto a chromatographic column.
The solvent program for elution was modified from a previously described method (20). This modification was made to increase the sensitivity of detection by increasing peak sharpness and resolving PG from other chemical compounds coextracted from brain tissue. The solvent system was composed of 0.1% formic acid in water (solvent A) and 0.1% formic acid in acetonitrile (solvent B). The flow rate was 0.2 ml/min, and the initial solvent conditions started with 10% solvent B. At 2 min, the percentage of B was increased to 65% over 8 min; at 15 min, the percentage of B was increased to 90% over 5 min; and at 35 min, it was reduced to 10% over 2 min. Equilibration time between runs was 13 min.
MS analysis was performed using a quadrupole mass spectrometer (API3000; Applied Biosystems, Foster City, CA) equipped with a TurboIonSpray ionization source. Analyst software version 1.4.2 (Applied Biosystems) was used for instrument control, data acquisition, and data analysis. The mass spectrometer was optimized in the multiple reaction monitoring mode. The source was operated in negative ion electrospray mode at 450jC, electrospray voltage was 24,250 V, nebulizer gas was zero grade air at 8 l/min, and curtain gas was ultrapure nitrogen at 11 l/min. Declustering potential, focusing potential, and entrance potential were optimized individually for each analyte as presented in Table 1. Focusing potential was 2200 V, and entrance potential was 210 V for all analytes. The quadrupole mass spectrometer was operated at unit resolution. PGE 2 , PGD 2 , PGF 2a , and TXB 2 were quantified using PGE 2 -d4 as the internal standard, whereas 6-oxo-PGF 1a was quantified using 6-oxo-PGF 1a -d4 as the internal standard. Initially, we used PGD 2 -d4, PGF 2a -d4, and TXB 2 -d4 to normalize PGD 2 , PGF 2a , and TXB 2 , respectively; however, this approach did not improve variability or recovery results compared with quantification using only PGE 2 -d4. An example of brain PG LC-ESI-MS/MS analysis is presented in Fig. 1.

Statistical analysis
All statistical comparisons were calculated using a two-way, unpaired Student's t-test or a one-way ANOVA and a Tukey-Kramer posthoc test when appropriate, using Instat II (Graphpad, San Diego, CA). Statistical significance was defined as P , 0.05. All values are expressed as means 6 SD.

PG extraction
The brain has a complicated biological matrix that produces high chemical background noise, so that the use of LC-MS/MS to measure brain PG mass requires special care with regard to sample preparation before analysis. Although several methods for brain PG extraction and purification are currently used (9,12,(14)(15)(16)21), these methods have not been evaluated for use by LC-MS/MS analysis.
Here, we have evaluated the background chemical noise and limits of sensitivity for four different methods currently used for PG extraction and analysis. The efficiency of the extraction procedures was estimated by the recovery of the deuterium-labeled PG added to the samples before extraction and ranged from 85% to 95% for each of the methods tested; there were no significant differences between methods.

Brain extraction with hexane/2-propanol
We also tested hexane/2-propanol extraction (3:2, v/v), which allows for simultaneous extraction of brain phospholipids, neutral lipids, and PG (14). Briefly, ?20 mg of brain microwaved tissue was homogenized in 1 ml of hexane/2-propanol (3:2, v/v) containing PGE 2 -d4 and 6-oxo-PGF 1a -d4 (100 pg in 10 ml of acetonitrile) as internal standards using a 2 ml Tenbroeck tissue grinder. The homogenizer was rinsed three times with 1 ml of hexane/ 2-propanol. The homogenate was subjected to 10 min of centrifugation (2,000 g) at 4jC, and the supernatant was removed. The supernatant was then dried under a stream of nitrogen, redissolved in 15% methanol at pH 3, and purified on C 18 cartridges as described above (10). PG were prepared for LC-MS/MS analysis as described in Materials and Methods. The chemical background noise was similar to values for methanol extraction followed by solid-phase extraction on C 18 columns. Also, the sensitivity was 4-to 16-fold lower compared with the acetone liquid/ liquid extraction described in Materials and Methods (Table 2).
Because purification of hexane/2-propanol lipid extracts on C 18 cartridges did not increase the levels of sensitivity of LC-MS/MS analysis compared with other methods, we used acetone to reextract PG from hexane/2-propanol lipid extracts. To purify hexane/2-propanol lipid extracts with acetone, aliquots of extracts were transferred into silanized tubes, solvent was removed under a stream of nitrogen and redissolved in 2 ml of acetone containing 0.005% BHT, and then 1 ml of saline was added. This mixture was mixed by vortexing for 4 min. We then followed the procedures for PG analysis as described in Materials and Methods. Although the purification of lipid extract with acetone followed by liquid/liquid extraction did not significantly improve the levels of sensitivity compared with direct extraction with acetone (Table 2), this approach allows the simultaneous analysis of PG and other lipid molecules in the same sample.

Extraction recovery with acetone
We verified the recovery of PG extraction with acetone compared with extraction with hexane/2-propanol, because hexane/2-propanol affords a high recovery of PG from tissue (14). Because we found that hexane/2propanol extraction has reduced sensitivity as a result of high background, we induced brain PG formation by injecting mice with lipopolysaccharide (1 mg/kg ip) (22) at 3 h before head-focused microwave irradiation. We extracted the same brain samples using either acetone or hexane/2-propanol (3:2, v/v). There were no differences in PG mass found after hexane/2-propanol extraction compared with acetone extraction (Fig. 2), indicating that acetone yields a high extraction ratio of PG from the brain tissue. Because acetone extraction produced significantly less background chemical noise compared with hexane/2-propanol extraction, the standard deviations for the individual PG are smaller in samples analyzed from the acetone extract compared with the hexane/2propanol extract.  Methanol followed by solid-phase extraction 3.9 6 1.0 (0. 5  In summary, tissue PG extraction with acetone followed by liquid/liquid extraction significantly increased the level of sensitivity in the LC-MS/MS analysis compared with other extraction methods tested in this study. This increased sensitivity was the result of a significant reduction of background chemical noise, probably attributable to better purification of PG extract from components that affect LC-MS/MS analysis. Also, dissolving the residue from a hexane/2-propanol lipid extract with acetone permits PG analysis in common lipid extracts that contain all major lipids, thereby extending the application of this method of extraction to other lipid parameters beyond PG, which is important when sample quantity is limited. Besides better sensitivity, PG extraction with acetone followed by liquid/liquid extraction is considerably less laborious when large sets of samples are analyzed and much less expensive compared with purification on C 18 cartridges. Therefore, we consider acetone extraction followed by liquid/liquid extraction to be the method of choice for PG analysis using LC-MS/MS.

Sample preparation for LC-MS/MS analysis
The solvent composition used to apply a sample onto an HPLC column may have a significant effect on analyte separation and peak sharpness, thereby affecting the limits of detection and the accuracy of analysis. Acetonitrile would be the best solvent to dissolve a sample before application on the column because it dissolves PG and is a component of the mobile phase. However, because of the small volume of the HPLC column used for PG separation and the low flow rate of solvents used in the LC separation program, a high acetonitrile concentration used to apply the sample reduces peak sharpness and increases peak leading (Fig. 3). We have found that the optimal acetonitrile-water ratio is 1:2 when 25 ml of sample is applied onto an HPLC column. Fig. 3. Effect of loading solvent composition of the separation of brain PG by HPLC. A sample of brain tissue (10 mg) from mice euthanized by decapitation was extracted and analyzed using a procedure described in Materials and Methods. The sample (25 ml) was loaded onto an HPLC column using either water-acetonitrile (1:2) (upper panel) or water-acetonitrile (2:1) (lower panel). cps, counts per second.  (100, 100, 100, 100, 50, and 50 pg, respectively, in 10 ml of acetonitrile) was analyzed using a procedure described in Materials and Methods. cps, counts per second.

Antioxidants prevent PG degradation during analysis
BHT is often used to prevent PG oxidation during extraction and analysis. Different concentrations of BHT have been used in lipid analysis, ranging from 0.1% to 0.01% (3,30), although not all investigators use antioxidants during PG analysis. To evaluate the need for antioxidants, we analyzed three identical brain samples with 0.1% BHT, 0.005% BHT, or no BHT added to acetone and chloroform used in the extraction. We found that 0.1% BHT produced a precipitate that may clog the LC system. However, using 0.005% BHT in chloroform and acetone for PG extraction and sample preparation efficiently decreased the variability of analysis and prevented a 2.8-fold reduction in 6-oxo-PGF 1a mass without producing a precipitate in the loading mixture.

Use of head-focused microwave irradiation
Brain PG mass found in rodents euthanized by headfocused microwave irradiation is 10-to 40-fold lower than in animals euthanized by decapitation (17,18). Different processes may account for this difference in results, including postmortem PG formation (17)(18)(19), induction of Fig. 5. Induction of PG formation during brain tissue extraction. Two groups of animals were anesthetized with halothane (1-3%) and euthanized by decapitation, and their skulls were opened. The whole brains were removed at 1 min after animal death and either frozen in liquid nitrogen (A) or subjected to head-focused microwave irradiation (B) as described in Materials and Methods. A: Three identical nonmicrowaved brain samples were extracted either in ice-cold conditions or after sample incubation in extraction mixture for 1 or 5 min at room temperature. B: As a control for the completeness of extraction in ice-cold conditions, microwaved brains were extracted under the same conditions described for A. Values are expressed as means 6 SD (n 5 3). * P , 0.05 compared with samples extracted in ice-cold conditions. ww, wet weight. PG formation during extraction, or PG destruction or trapping in microwaved brains.
Because the induction of PG formation during extraction has not been tested in previous studies, we extracted identical nonmicrowaved brain samples either in ice-cold conditions or after sample incubation in extraction mixture for 1 or 5 min at room temperature. The mass of all PG analyzed was increased significantly at room temperature (Fig. 5A). As a control for the completeness of extraction in ice-cold conditions, microwaved brains with heat-inactivated enzymes were extracted under the same conditions described above. There were no differences between PG mass in microwaved tissue after extraction in ice-cold conditions compared with room temperature (Fig. 5B), indicating the same high extraction ratio in icecold conditions. These data indicate that induction of PG formation during extraction can occur, which is one factor contributing to the higher values and greater variability of PG mass found in nonmicrowaved brains.
Another reason for a reduction in PG mass in brains subjected to microwaved irradiation could be PG breakdown and/or trapping in irradiated tissue. Temperatures of 70jC to 100jC are reported a few seconds after expos-ing the animal to head-focused microwave irradiation (19), similar to our observations. Because PG are known for their instability and have a short half-life, PG breakdown in heat-denatured tissue should be considered as another factor accounting for the reduced PG mass observed in microwaved brains compared with nonmicrowaved brains. In addition, because PG are bound in vivo by a variety of carrier proteins (31,32), trapping of PG in heatdenatured proteins may also account for the reduction in PG mass observed after microwave irradiation. Although it is assumed that the observed reduction in PG mass in microwaved versus nonmicrowaved brain is the result of the heat inactivation of enzymes involved in postmortem PG formation (17)(18)(19), direct evidence that PG are not trapped or destroyed in microwaved brains has not been reported. What has been reported is that in indomethacintreated rats euthanized by decapitation, brain PG mass is not different from the levels in brains from rats euthanized by microwave irradiation (18). Additional evidence is that after injection (intracerebral ventricular) of radiolabeled PG into the brain before microwave irradiation, most of the recovered radioactivity was in the form of PG (19), suggesting heat stability in situ. However, the recovery of Fig. 6. PG are not destroyed or trapped in microwaved brain. Two groups of animals were anesthetized with halothane (1-3%) and euthanized by decapitation, and their skulls were opened. The whole brains were removed at 1 min after animal death and either frozen in liquid nitrogen or subjected to head-focused microwave irradiation as described in Materials and Methods. The temperature of microwaved brains ranged from 70jC to 85jC as measured using a thermocouple. As a control for enzyme inactivation in microwaved brains, a third group of animals was euthanized by microwave irradiation as described in Materials and Methods. Values are expressed as means 6 SD (n 5 3). * P , 0.05 compared with brains from mice euthanized by head-focused microwave irradiation without global ischemia; ** P , 0.05 compared with brains from nonmicrowaved brains of mice exposed to global ischemia. ww, wet weight. Fig. 7. Reduction in PG mass during tissue storage. Identical tissue samples were analyzed either the same day or 4 weeks after animals were euthanized by decapitation. Tissue powder was stored at 280jC. Values are expressed as means 6 SD (n 5 4). * P , 0.05 compared with brains analyzed the same day as the mice were euthanized. the radiolabeled PG from brains subjected to microwave irradiation was not examined in this study.
To address the possibility that microwave irradiation may affect the recovery of endogenous PG, we induced brain PG production by modeling global ischemia. Two groups of animals were anesthetized with halothane (1-3%) and euthanized by decapitation, and their skulls were opened. The whole brains were removed 1 min after death and either frozen in liquid nitrogen or subjected to microwave irradiation as described in Materials and Methods. The temperature of microwaved brains ranged from 70jC to 85jC as measured using a thermocouple. As a control for enzyme inactivation in microwaved brains, the third group of animals was euthanized immediately using headfocused microwave irradiation as described in Materials and Methods. The magnitude of increased PG mass found in nonmicrowaved brains was similar to the levels reported by others (17,18) (Fig. 6). For most PG analyzed, there were no differences between the two different fixation regimens subjected to induction of PG production via ischemia (Fig. 6), indicating that the recovery of brain PG was fixation-independent.
However, PGD 2 and TXB 2 mass were increased in nonmicrowaved brains compared with microwaved brains after induction of PG formation. There are several explanations for these data. First, PGD2 and TXB2 are formed during extraction because their mass was affected to a greater extent than the mass of other PG during the extraction of nonmicrowaved brains (Fig. 5A). Second, PGD 2 and TXB 2 are more heat-labile. To test this assumption, we incubated a PGE 2 , PGD 2 , PGF 2a , TXB 2 , and 6-oxo-PGF 1a mixture dissolved in acetonitrile-water (1:2, v/v) at 85jC for 10 min and analyzed PG mass as described in Materials and Methods. The mass of all PG tested was decreased to the same extent (10-15%; data not shown), indicating similar heat lability of the tested PG.
Together, these data support the need to microwave brain and provide direct evidence that PG are not trapped or destroyed in microwaved brains.

PG stability during tissue storage
Because of PG short shelf half-lives, we tested PG stability in brain tissue and extracts during storage. PG were analyzed in the same samples either the same day or 4 weeks after animals were euthanized by decapitation. Tissue powder was stored at 280jC. Four weeks of tissue powder storage at 280jC resulted in a 2-to 4-fold decrease in PG mass (Fig. 7); however, no decrease of PG was observed in lipid extracts stored in hexane/2-propanol (3:2) for several months at 280jC under nitrogen (data not shown). These data indicate the utility of rapid tissue extraction with hexane/2-propanol (3:2), which then can be stably stored in hexane/2-propanol for PG and other lipid analysis in the future.

Summary
In summary, tissue PG extraction with acetone followed by liquid/liquid extraction significantly increased the level of sensitivity of LC-MS/MS analysis compared with other extraction methods tested in this study. This increased sensitivity was the result of a significant reduction in background chemical noise. Dissolving residue from hexane/ 2-propanol lipid extracts with acetone allows the analysis of PG in these lipid extracts, thereby extending the application of this method of extraction. Besides better sensitivity, this method is less laborious and less expensive compared with purification on C 18 cartridges. We also evaluated PG stability during extraction and storage. The use of 0.005% BHT during PG extraction decreased the variability of analysis and limited 6-oxo-PGF 1a oxidation. Importantly, PG were rapidly destroyed during the storage of powdered tissue; however, PG were stable in hexane/2-propanol extracts. Lastly, our data support the need to euthanize animals by head-focused microwave irradiation rather than by decapitation and provide direct evidence that PG are not trapped or destroyed in microwaved brains.