Article pubs.acs.org/crt
Cite This: Chem. Res. Toxicol. 2018, 31, 176−182
Analysis of Urinary Eicosanoids by LC−MS/MS Reveals Alterations in the Metabolic Profile after Smoking Cessation Michael Goettel,†,‡ Reinhard Niessner,† Max Scherer,‡ Gerhard Scherer,‡ and Nikola Pluym*,‡ †
Chair for Analytical Chemistry, Technische Universität München, Marchioninistraße 17, 81377 Munich, Germany ABF, Analytisch-Biologisches Forschungslabor GmbH, Semmelweisstraße 5, 82152 Planegg, Germany
‡
ABSTRACT: A preceding untargeted metabolic fingerprinting approach in our lab followed by targeted fatty acid analysis revealed alterations in arachidonic acid metabolism in samples derived from a diet-controlled smoking cessation study in which compliant subjects (N = 39) quit smoking at baseline and were followed over the course of 3 months. Consequently, urinary eicosanoids were evaluated by means of a validated LC−MS/MS method. A significant decrease was obtained for the prostaglandins PGF2α, 8-iso-PGF2α, thromboxane 2,3-dTXB2, and leukotriene E4 upon quitting smoking. These findings indicate a partial recovery of smoking-induced alterations in the eicosanoid profile due to a reduction in oxidative stress and the inflammatory response.
■
INTRODUCTION Untargeted metabolic fingerprinting in combination with targeted analysis is a powerful tool to elucidate alterations between distinctive groups. Such a combined approach of untargeted profiling and targeted quantification has been performed recently in our laboratory in an initial clinical study1,2 to identify differences in the metabolome of smokers and nonsmokers. We found several major metabolic pathways that differ significantly between these groups, especially lipid and fatty acid metabolism.1−3 The findings from the initial clinical study comparing smokers and nonsmokers encouraged us to conduct a second diet-controlled clinical study in which we investigated metabolic alterations in smokers after cessation. Untargeted fingerprinting analysis using a validated GC−TOF−MS method1,2 showed fatty acid (FA), amino acid, and energy metabolism changes, indicating a partial recovery of the smokers’ metabolome toward the nonsmoker profile upon 3 months of cessation.4 Subsequently, we developed and validated a targeted GC− TOF−MS method for the quantification of FAs.5 Several FA species were significantly altered after smoking cessation, including saturated as well as mono- and polyunsaturated FAs. Among other FAs, arachidonic acid showed a statistically significant increase after 1 month of smoking cessation, which prompted us to investigate the eicosanoid profile in this second clinical study as arachidonic acid (FA 20:4 5Z,8Z,11Z,14Z) is regarded as a key precursor toward the formation of various eicosanoids. Eicosanoids are signaling molecules produced through both enzymatic and nonenzymatic free radical catalyzed reactions.6 They are secreted from different human cell types under normal and pathophysiological conditions. Alterations in © 2018 American Chemical Society
eicosanoid levels may be caused by various disease states like asthma,7,8 diabetes,9 and cancer9−11 as well as exogenous sources like smoking.9,12,13 Hence, eicosanoids are often investigated in plasma and urine as biomarkers indicating oxidative stress or inflammation.6,10,12,14,15 Therefore, we decided to analyze urinary eicosanoids using a validated, targeted LC−MS/MS method.12 In terms of smoking cessation, only few analytes were investigated simultaneously in previous studies.16,17 Thus, we decided to extend our method to a total of nine eicosanoids by inclusion of two additional analytes, namely, the prostaglandins (PGs) tetranor-PG D-metabolite (tPGDM) and PGF2α, to allow for the investigation of a more comprehensive eicosanoid profile (Figure 1).
■
MATERIALS AND METHODS
Reagents and Standards. Acetic acid (≥99%), ammonium hydroxide (28% in water), creatinine (anhydrous), formic acid (≥95%), hydrochloric acid (∼37%), and sodium hydroxide (≥97%, pellets) were purchased from Sigma-Aldrich (Munich, Germany). Chloroform (picograde), ethyl acetate (optigrade), methanol (optigrade), and water (optigrade) were obtained from LGC Standards (Wesel, Germany). Tetranor PGE-M (t-PGEM), tetranor PGD-M (tPGDM), 2,3-dinor-8-iso-PGF2α (2,3-d-8-iso-PGF2α), 8-iso-PGF2α, 2,3dinor-TXB2 (2,3-d-TXB2), 11-dehydro-TXB2 (11-dh-TXB2), LTE4, 12(S)-HETE, D6-tetranor-PGDM, D6-tetranor-PGEM, D4-8-isoPGF2α, D4-11-dehydro-TXB2, D5-LTE4, and D8-12(S)-HETE had purities higher than 97% and were purchased from Biomol (Hamburg, Germany). 0.1% formic acid in water (ULC−MS grade) and 0.1% formic acid in acetonitrile (ULC−MS grade) were from Biosolve BV (Valkenswaard, Netherlands). Received: October 6, 2017 Published: February 5, 2018 176
DOI: 10.1021/acs.chemrestox.7b00276 Chem. Res. Toxicol. 2018, 31, 176−182
Article
Chemical Research in Toxicology
Figure 1. Structures of the nine eicosanoids analyzed by LC−MS/MS. Smoking Cessation Study and Sample Collection. The clinical study was carried out as described in detail in Goettel et al.4 The protocol was approved in accordance with the Helsinki declaration18 by the ethical commission of the Medical Chamber of North RhineWestphalia, Germany. Briefly, 39 smokers (15−24 cigarettes/day) were followed over the course of 3 months of smoking cessation. Subjects were healthy male individuals who were still smoking at the start of the study (TP0). Sampling was conducted during the inpatient stays at the start of the study (TP0) and after 1 week (TP1), 1 month (TP2), and 3 months (TP3) of smoking cessation. A controlled diet with 72% carbohydrates, 14% fat, and 14% protein was necessary in order to reduce nutritional effects on eicosanoid synthesis and metabolism. Hence, 12 hours of fasting prior to the in-patient stays were mandatory. Meals were served at defined points during confinement (in-patient stays at TP0−TP3). Moreover, the caloric intake was adjusted according to the individuals’ body weight (55−65 kg: ∼7500 kJ; 66−75 kg: ∼8500 kJ; 76−85 kg: ∼9400 kJ; 86−95 kg: ∼10 200 kJ; and 96−105 kg: 10 700 kJ). Compliance was verified by determination of carbon monoxide in exhaled breath and cotinine in saliva and urine in several ambulatory visits interspersed throughout the entire clinical study. Subjects with COex levels above 6 ppm and/or concentrations of cotinine over 15 ng/mL in saliva and 50 ng/mL in urine, respectively, were regarded as noncompliant and excluded from the clinical study. In total, 39 participants of the initially 60 subjects were compliant over the complete course of 3 months and selected for urinary bioanalysis of eicosanoids and subsequent data evaluation. Urine samples were collected in three fractions at the in-patient stays (TP0−TP3) under strictly controlled conditions during a period of 24 h of controlled diet (8 am first day to 8 am second day). For the analysis of eicosanoids, the urine fractions were combined proportionally to yield a 24 h urine sample. Samples were stored at −20 °C until analysis. Quality Control Samples. In addition to the set of study samples, nine quality control samples at low, medium, and high concentration levels (three per level) were interspersed throughout the entire analytical batch. Therefore, urine samples were pooled and partly spiked or diluted to yield concentrations in the low, medium, and high calibration ranges. The acceptance criteria for the accuracy of QC samples were set according to FDA guidelines19 at ±20% for the low level and ±15% for the medium and high levels. At least two-thirds of all QC samples and 50% per level had to meet the acceptance criteria; otherwise, the batch had to be repeated. Method Validation. In addition to the seven analytes implemented initially [tetranor PG E-metabolite (t-PGEM), 8-isoand 2,3-dinor-8-iso-PGF2α (8-iso-PGF2α; 2,3-d-8-iso-PGF2α), the thromboxanes (TX) 11-dehydro- and 2,3-dinor-TXB2 (11-dh-TXB2,
2,3-d-TXB2), leukotriene (LT) E4 (LTE4), and 12-hydroxyeicosatetraenoic acid (12-HETE)], two further analytes were introduced, namely, PG D-metabolite (t-PGDM) and PGF2α, to include additional established inflammation and oxidative stress markers and expand the method’s profile to a total of nine eicosanoids. The method was revalidated with respect to the novel analytes t-PGDM and PGF2α in analogy to Sterz et al.12 and according to FDA guidelines.19 Calibration. Quantification was performed with the internal standard method. Separate calibrations were performed with a set of 8−10 calibrators for each analyte, depending on the target eicosanoid and the calibration range as described in Sterz et al.12 (for data of the additional analytes, cf. Table 2). Calibrators were prepared by spiking nonsmoker urine samples with increasing amounts of analytes. For the calibration of the new analytes, [D6]-t-PGDM was used as an internal standard (IS) for t-PGDM and [D4]-11-dh-TXB2 was used for PGF2α. Evaluation of the chromatographic data was performed using Analyst 1.6.3 (AB Sciex, Darmstadt, Germany) and Excel 2013 (Microsoft, Redmond, USA). Calculation of the calibration curves was done by linear regression with 1/× weighting. Sample Preparation. Urine samples from the clinical study4 were prepared and measured according to Sterz et al.12 All samples were randomized prior to LC−MS/MS analysis. Aliquots of 3 mL from the 24 h urine samples were used for analysis. According to Sterz et al.,12 20 μL of acetic acid and 30 μL of an IS mixture containing 6 ng of [D6]-t-PGEM, 6 ng of [D6]-t-PGDM, 6 ng of [D4]-8-iso- PGF2α, 6 ng of [D4]-11-dh-TXB2, 1.5 ng of [D5]-LTE4, and 1.5 ng of [D8]-12HETE were added to each sample prior to extraction. Extraction was performed according to Bligh and Dyer20 with modifications. For liquid−liquid extraction, 11.25 mL of B&D solution (methanol/chloroform 2:1 v/v) was added to each sample. After vortexing, samples were incubated for 1 h at ambient temperature. Subsequently, 3.75 mL of chloroform and 3.75 mL of water were added. The samples were vortexed again and centrifuged for 10 min at 2500 rpm. The bottom layer chloroform phase was transferred into a vial and evaporated to dryness in a SpeedVac centrifuge (Thermo Scientific, Dreieich, Germany). Finally, the residue was reconstituted in 100 μL of methanol for injection into the LC−MS/MS system. LC−MS/MS Analysis. An API 5000 triple quadrupole mass spectrometer (AB Sciex, Darmstadt, Germany) LC−MS/MS system, equipped with a 1200 series binary pump (G1312B), a degasser (G1379B), and a column oven (G1316B) (Agilent, Waldbronn, Germany) connected to an HTC Pal autosampler (CTC Analytics, Zwingen, Switzerland), was used for chromatographic separation. Negative electrospray ionization (ESI−) was performed on a Turbo V ion spray source (AB Sciex, Darmstadt, Germany). High-purity nitrogen was provided by a NGM 22-LC/MS nitrogen generator (cmc 177
DOI: 10.1021/acs.chemrestox.7b00276 Chem. Res. Toxicol. 2018, 31, 176−182
Article
Chemical Research in Toxicology
Figure 2. Representative UPLC−MS/MS chromatograms of a smoker’s urine after 3 months of cessation (TP3). Quantifier MRM transitions are given in the respective chromatograms. Numbers are assigned according to Table 3. Data Analysis. Different statistical tests were performed applying R (version 3.3.0) and RStudio (version 0.99.902) in order to follow alterations in the eicosanoid profile over the course of 3 months of smoking cessation. The Quade test was utilized in order to identify significant (α = 0.05) changes across all points in time. The Wilcoxon signed rank test was applied for cessation-relevant comparisons between two points in time (e.g., TP0/TP1) in order to verify the results of the Quade test. To counteract the problem of multiple statistical comparisons, the significance levels were adjusted by Bonferroni correction.21,22
Instruments, Eschborn, Germany). For chromatographic separation, a Waters (Eschborn, Germany) Acquity ultra performance liquid chromatography (UPLC) BEH C18 column (2.1 × 50 mm) with 1.7 μm particle size was utilized. The column oven was operated at 30 °C. Five microliters of a sample were injected at a constant flow of 600 μL/min. Eluents were 0.1% formic acid in water (A) and 0.1% formic acid in acetonitrile (B). Separation of the analytes was achieved with gradient elution starting with 5% B for 1 min followed by a linear increase to 53% B until 9.5 min and a second linear increase to 76% B until 11 min. Third, a step to 100% B until 11.1 min was performed and held for 1 min. Finally, re-equilibration of the analytical system was achieved by rinsing from 12.1 to 14 min with 5% B. The turbo ion spray source was operated with the following settings: ion spray voltage = −4 kV, heater temperature = 600 °C, source gas 1 = 20 psi, source gas 2 = 5 psi, CAD gas = 5 psi, and curtain gas = 40 psi. Data acquisition was performed with quadrupoles working at unit resolution and multiple reaction monitoring (MRM).
■
RESULTS AND DISCUSSION The determination of nine individual eicosanoid species was performed by LC−MS/MS operating in ESI-negative mode. Baseline separation of the isobaric species PGF2α/8-iso-PGF2α and t-PGDM/t-PGEM was achieved within a total run time of 14 min (Figure 2). Gradient elution yielded satisfactory peak 178
DOI: 10.1021/acs.chemrestox.7b00276 Chem. Res. Toxicol. 2018, 31, 176−182
Article
Chemical Research in Toxicology Table 1. Precision, Accuracy, Recovery, and Stability Investigations for t-PGDM and PGF2α conc.
intraday precision (N = 5)
conc.
accuracy (N = 5)
recovery
conc.
short-term stability r.t. after 24 h
analyte
[ng/mL]
CV [%]
CV [%]
[ng/mL]
[%]
[%]
[ng/mL]
[%]
[%]
[%]
t-PGDM
0.40 1.00 10.00 0.40 1.10 9.20
11.0 5.8 3.8 4.5 6.7 3.8
5.2 7.2 2.8 6.0 5.4 3.6
0.25 1.00 10.00 0.25 1.00 10.00
93.5 94.3 92.7 94.2 96.5 93.4
23.6 20.5 20.3 66.7 72.7 68.5
0.40
89.8
89.7
96.3
10.90 0.80
91.0 112.0
101.0 98.3
105.4 112.1
9.00
106.0
106.0
110.5
PGF2α
interday precision (N = 6)
freeze/thaw stability after 6 cycles
long-term stability at
