Analyst View Article Online
Published on 31 July 2014. Downloaded by Christian Albrechts Universitat zu Kiel on 28/10/2014 12:14:23.
PAPER
View Journal | View Issue
Cite this: Analyst, 2014, 139, 5600
Rapid assessment of the coenzyme Q10 redox state using ultrahigh performance liquid chromatography tandem mass spectrometry† Zhi Tang,a Shangfu Li,a Xinyuan Guan,b Philippe Schmitt-Kopplin,c Shuhai Lin*a and Zongwei Cai*a An improved method for accurate and rapid assessment of the coenzyme Q10 (CoQ10) redox state using ultrahigh performance liquid chromatography tandem mass spectrometry was described, with particular attention given to the instability of the reduced form of CoQ10 during sample preparation, chromatographic separation and mass spectrometric detection. As highly lipophilic compounds in complex biological matrices, both reduced and oxidized forms of CoQ10 were extracted simultaneously from the tissue samples by methanol which is superior to ethanol and isopropanol. After centrifugation, the supernatants were immediately separated on a C18 column with isocratic elution using methanol containing 2 mM ammonium acetate as a non-aqueous mobile phase, and detected by positive electrospray ionization tandem mass spectrometry in multiple reaction monitoring (MRM) mode. Ammonium acetate as an additive in methanol provided enhanced mass spectrometric responses for both forms of CoQ10, primarily due to stable formation of adduct ions [M + NH4]+, which served as precursor ions in positive ionization MRM transitions. The assay showed a linear range of 8.6–8585 ng mL1 for CoQ10H2 and 8.6–4292 ng mL1 for CoQ10. The limits of detection (LODs) were 7.0 and 1.0 ng mL1 and limits of quantification (LOQs) were 15.0 and 5.0 ng mL1 for CoQ10H2 and CoQ10, respectively. This rapid
Received 29th April 2014 Accepted 31st July 2014
extractive and analytical method could avoid artificial auto-oxidation of the reduced form of CoQ10, enabling the native redox state assessment. This reliable method was also successfully applied for the
DOI: 10.1039/c4an00760c
measurement of the CoQ10 redox state in liver tissues of mice exposed to 2,3,7,8-tetrachlorodibenzo-p-
www.rsc.org/analyst
dioxin, revealing the down-regulated mitochondrial electron transport chain.
1
Introduction
Endogenous coenzyme Q10 (CoQ10) is a highly lipophilic small molecule consisting of a benzoquinone head group and 10 isoprenyl subunits in its side chain, which coexists in two different types – the reduced form (CoQ10H2, ubiquinol) and the oxidized form (CoQ10, ubiquinone) (Fig. 1), which is present mainly in the mitochondria of eukaryotic cells. As a crucial component in the mitochondrial respiratory chain for electron transport and cellular energy production,1 CoQ10 plays an important role in a wide range of biochemical, physiological and pathological processes. For example, it affects gene expression involved in cell metabolism,2,3 and reduced levels of
CoQ10 are implicated in different diseases, like breast cancer,4 diabetes,5 and neurodegenerative diseases.6 Given the vital role of CoQ10 in biological functions, its use as a nutritional supplement or adjuvant therapy for various diseases, including cancer,7 cardiovascular disease,8 and diabetes9 and its possible effect of limiting the side effects of statin therapy,10 have also been reported in recent investigations. It has become obvious that the unique antioxidant and health-promoting properties of CoQ10 are largely dependent both on its concentration and
a
State Key Laboratory of Environmental and Biological Analysis, Department of Chemistry, Hong Kong Baptist University, Hong Kong SAR, China. E-mail: zwcai@ hkbu.edu.hk; [email protected]; Fax: +852 3411 7348; Tel: +852 3411 7070
b
Department of Clinical Oncology, University of Hong Kong, Hong Kong SAR, China
c
Analytical Biogeochemistry, Helmholtz-Zentrum Muenchen–German Research Center for Environmental Health, Ingolstaedter Landstrasse 1, D-85764, Oberschleißheim, Germany † Electronic supplementary 10.1039/c4an00760c
information
5600 | Analyst, 2014, 139, 5600–5604
(ESI)
available.
See
DOI:
Fig. 1 Reduced form of CoQ10 can be easily oxidized to its corresponding oxidized form (left: reduced form, CoQ10H2, ubiquinol; right: oxidized form, CoQ10, ubiquinone).
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 31 July 2014. Downloaded by Christian Albrechts Universitat zu Kiel on 28/10/2014 12:14:23.
Paper
redox state. To advance the fundamental understanding of the roles of CoQ10 in various physiological and pathological conditions, as well as an efficiency evaluation of its supplementation, an analytical method for assessing the CoQ10 content and its redox state in complex biological matrices is required. Therefore, the aim of the current study was to develop a specic and sensitive mass spectrometry-based method for rapid quantication of CoQ10 in biological samples. In previous studies, CoQ10 concentrations in different biological specimens, including serum, cells, subcellular organelles, or other clinical and research samples were analyzed by high performance liquid chromatography (HPLC) coupled with a tandem mass spectrometer11–15 or an electrochemical detector.16–18 Unlike HPLC coupled with electrochemical detection, ultrahigh performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) would be a newly developed method with high sensitivity and high selectivity to measure both forms of CoQ10. However, due to the instability of the reduced form of CoQ10 which can be easily oxidized to its corresponding oxidized form during sample pre-treatment,19–21 reliable assay of the coenzyme should be further explored. With particular attention to this problem, here, we described a rapid quantitative assay for characterization of the CoQ10 redox state in mouse liver tissue samples using UPLC-MS/MS.
2 Experimental 2.1
Chemicals and reagents
2,3,7,8-Tetrachlorodibenzo-p-dioxin (abbr. TCDD, purity >99%) was obtained from AccuStandard, Inc. (New Haven, CT, USA). Coenzyme Q10 standard (oxidized form, >98%, HPLC), and sodium borohydride (NaBH4) were obtained from SigmaAldrich (St. Louis, MO, USA) and used without further purication. Dipropoxy-CoQ10 (DP-Q10), a propyl analogue of CoQ10 served as the internal standard, was prepared according to a recently reported protocol.22 Methanol, ethanol, isopropanol, and hexane were of HPLC grade and were obtained from Tedia (Faireld, OH, USA). Ultrapure water was puried by using a Milli-Q Academic Water Purication System from Millipore (Millipore, Bedford, MA, USA). 2.2
Instrumentation and UPLC-MS/MS analysis
LC-MS/MS analyses were performed on a Waters ACQUITY ultrahigh performance liquid chromatograph (UPLC) system hyphenated with a TQD triple quadrupole mass spectrometer tted with a Z-spray electrospray ionization source. The chromatographic separation was achieved by using a C18 reverse phase column (ACQUITY UPLC BEH C18 column, 2.1 mm 50 mm, 1.7 mm) employing methanol containing 2 mM ammonium acetate as the mobile phase at the ow rate of 0.4 mL min1 under isocratic elution conditions. The column temperature was kept at 40 C. The mass spectrometer was operated in positive ionization mode. The optimized parameters were capillary voltage, 2 kV; cone voltage, 30 V; source and desolvation temperatures were 120 C and 400 C, respectively. Nitrogen gas was used as the desolvation gas (700 L h1) and
This journal is © The Royal Society of Chemistry 2014
Analyst
cone gas (50 L h1). Argon gas was used for collision-induced dissociation. Multiple reaction monitoring (MRM) was used for quantication of both reduced and oxidized CoQ10. MRM transitions, cone voltages and collision energies as well as dwell times are summarized in Table 1. MassLynx v.4.1 soware was used for data acquisition and analysis (Waters Corp., Milford, MA, USA). 2.3
Animal experiment
Eight week old male C57BL/6J mice were purchased from the Chinese University of Hong Kong. The animal experimental procedures were approved by the Committee on the Use of Live Animals for Teaching and Research, Department of Health, Hong Kong SAR, China and were carried out in accordance with the NIH Guide for the Care and Use of Laboratory Animals. The mice were divided into control groups (n ¼ 7) and TCDD treated groups (n ¼ 7 for the high dosage group; n ¼ 7 for the low dosage group). Mice in each group were treated with either corn oil (the control group) or TCDD dissolved in corn oil (2.0 mg mL1 and 0.2 mg mL1) for three consecutive days by oral administration of 20 mg kg1 and 2.0 mg kg1 body weight, respectively. On the tenth day, all the mice were decapitated, and the liver tissues were harvested, snap frozen with liquid nitrogen and then stored at 80 C. 2.4
Standard solution preparation
One milligram per milliliter CoQ10 stock solution was prepared as follows: 25 mg of CoQ10 standard was accurately weighed and placed into a 25 mL volumetric ask. Three milliliters hexane as the co-solvent was added to dissolve the solid completely, and the resultant solution was diluted to the mark with isopropanol and stored at 20 C. Working solutions of CoQ10 were prepared by diluting the stock solution with nitrogen-purged methanol. As the reference standard of CoQ10H2 (the reduced form of CoQ10) was not commercially available, its solutions were freshly prepared from the corresponding oxidized form upon reduction with sodium borohydride, as described by Frei et al.23 Briey, 200 mL of the CoQ10 stock solution was placed into an Eppendorf tube and 1.0 mg of solid NaBH4 was added. The mixture was then vortexed at room temperature. CoQ10 was converted quantitatively to CoQ10H2, as indicated by disappearance of ubiquinone and the appearance of product monitored by UPLC-MS/MS. Because of the instability of CoQ10H2, its solutions were prepared immediately prior to the analysis, and freshly prepared each time, and the working standard solutions
MRM transitions used for quantification of the reduced form and oxidized form of coenzyme Q10 and internal standard DP-Q10
Table 1
Analytes
MRM transition (m/z)
Dwell time (s)
CV (V)
CE (eV)
CoQ10H2 CoQ10 DP-Q10
883.0 / 197.0 881.0 / 197.0 937.0 / 253.0
0.1 0.1 0.1
30 30 30
25 25 20
Analyst, 2014, 139, 5600–5604 | 5601
View Article Online
Analyst
Paper
were generated by subsequent serial dilutions using deoxygenated methanol.
Published on 31 July 2014. Downloaded by Christian Albrechts Universitat zu Kiel on 28/10/2014 12:14:23.
2.5
Tissue sample pre-treatment and CoQ10 extraction
To avoid articial oxidation of the analytes, all solvents were deoxygenated shortly before use by purging with dry nitrogen and kept on ice. Three different alcoholic solvents, including methanol, ethanol, and isopropanol were tested to obtain optimal extraction efficiency. Each tissue was freeze-dried, pulverized, and thoroughly mixed, and then 30 mg freeze-dried powders were extracted with 300 mL ice-chilled alcoholic solvent containing 30 ng internal standard. The mixture was vortexed for 30 seconds and placed on ice for 15 min and then centrifuged at 12 000 rpm at 4 C for 10 min. This process was repeated twice; the supernatants were pooled together, then transferred into an auto-sampler vial and analyzed immediately. The remaining sample was stored at 20 C for a reproducibility study. 2.6
Method validation
Fig. 2 The full scan mass spectrum of oxidized CoQ10 (A) and product ion mass spectra of [M + NH4]+ (B).
Method validation was carried out in terms of precision, linearity, limits of detection, limits of quantitation, matrix effects, recovery and stability. Calibration curves were built using serial dilutions of a 1.0 mg mL1 CoQ10 standard solution, spiked with in-house made internal standard DP-Q10. The concentration levels of CoQ10 spanned from below 10 ng mL1 to near 10 000 ng mL1 while the concentration of the internal standard was maintained constant at 100 ng mL1. Three replicate injections were made for each calibration level.
aqueous reversed-phase mobile phase for chromatographic separation of CoQ10 is not only readily compatible with electrospray ionization mass spectrometry but also with tissue sample preparation. To achieve a very short run time in UPLC analysis as well as reproducible results, the column temperature was controlled, and temperatures ranging from 25 C to 40 C at an interval of ve degree were tested. As shorter retention times at higher column temperature were obtained, the column was maintained at 40 C during the following experiments (Fig. 4).
3 Results and discussion
3.2
3.1
UPLC-ESI-MS/MS method development
An improved UPLC-ESI-MS/MS method for accurate and rapid assessment of CoQ10 in mouse liver tissue samples was developed and validated, with particular attention given to chemical instability of CoQ10H2, and the low biological concentration levels present in complex matrixes. 3.1.1 ESI-MS/MS optimization. The MS/MS parameters were optimized by direct injection using the standard compounds. The most intense fragment ion in the product ion spectrum for each standard was selected to be monitored in MRM mode. The cone voltage (CV) and collision energy (CE) were optimized for MRM transitions of each [M + NH4]+ ion, which were predominantly observed in the positive ESI spectra (Fig. 2). The MRM transitions, cone voltages, collision energies, and internal standards are listed in Table 1. 3.1.2 UPLC separation optimization. The UPLC method was optimized to ensure resolution between the two types of coenzyme Q10. Different organic solvents, including methanol and acetonitrile as well as their combination were tested, and it was found that using methanol as the mobile phase provided well-resolved peaks with good peak shapes over the acetonitrile– methanol (1 : 4, v/v) solvent system (Fig. 3). The addition of 2 mM ammonium acetate as a mobile phase modier provided favorable sensitivities as [M + NH4]+ ions were used as parent ions in MRM mode. To be duly noted, methanol as a non-
5602 | Analyst, 2014, 139, 5600–5604
Coenzyme Q10 extraction optimization
Sample preparation is one of the major factors affecting the quantitative estimation of CoQ10 in biological samples. As reported previously,24,25 the use of alcohol for CoQ10 extraction combined with in situ chemical oxidation and followed by direct HPLC analysis is a fast and robust method for evaluating total CoQ10 in plasma. In this work, the efficiency of different alcohols, including methanol, ethanol, and isopropanol for the extraction of both forms of CoQ10, was evaluated. Methanol was found to be the most efficient extraction solvent for both forms
Fig. 3 Mobile phase composition on separation of the two forms of CoQ10 (A) methanol as the mobile phase results in well-resolved peaks; (B) acetonitrile–methanol (1 : 4, v/v) provides no separation.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 31 July 2014. Downloaded by Christian Albrechts Universitat zu Kiel on 28/10/2014 12:14:23.
Paper
Analyst
both forms of CoQ10 was achieved as indicated by imprecision less than 10% relative standard deviation. The relatively lower but consistent recovery (79%) for the spiked CoQ10 sample at the concentration of 45 ng mL1 (Table S2†) could be in part due to adsorption losses during sample preparation (for example, binding to biological matrices or being adsorbed on sample containers such as tubes or injection vials), but it is still practically acceptable. Calibration curves for CoQ10H2 and CoQ10 were linear over the range of 8.6 to 8585 ng mL1 and 8.6 to 4292 ng mL1 (Fig. S1 & S2†). The limits of detection (LODs) were 7.0 and 1.0, and the limits of quantitation (LOQs) were 15.0 and 5.0 ng mL1 for CoQ10H2 and CoQ10, respectively (Table S2†). More detailed results of the method performance and validation are provided in the ESI (Table S1, Fig. S3 and S4†).
3.4
Fig. 4 Column temperature on retention of both forms of CoQ10 and the internal standard (left to right: the 1st peak, CoQ10H2; the 2nd one, CoQ10; the last one, the internal standard DP-Q10).
of CoQ10 in our testing (Fig. 5). Thus, the optimized sample preparation method is single-step extraction of CoQ10 using methanol, which is compatible with chromatographic separation and mass spectrometric detection, eliminating liquid– liquid extraction and further concentration steps.
3.3
Method validation
Biological application
Previous studies have revealed that TCDD damaged the mouse liver in a dose-dependent manner.27,28 Abnormal levels of serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) activities were reported in TCDD-exposed mice. However, the underlying mechanisms of TCDD-induced hepatotoxicity in mice and humans are still not fully understood. The redox state in the liver may also reveal the toxic effect of TCDD exposure in mice. To measure whether the CoQ10 redox state was altered in mice exposed to TCDD, we applied the validated method to liver tissue samples, obtained from three different experimental mouse groups that were treated with low-dosage (LL), highdosage TCDD (HL), and control (CL). As shown in Fig. 6A, the amount of total CoQ10 (sum of the reduced form and oxidized
The developed method was validated in terms of precision, sensitivity, matrix effects, recovery, linearity, limit of detection (LOD), and limit of quantitation (LOQ) according to general guidelines for validation of the bioanalytical method from US Food and Drug Administration (FDA).26 Good reproducibility (precision at n ¼ 9, three parallel samples at low, medium, and high concentration levels, and each measured three times) of
Altered CoQ10 redox state with significant changes in liver tissues of mice exposed to TCDD. (A) Significantly decreased levels of the total CoQ10 in mice exposed to TCDD (CL vs. LL, and CL vs. HL, p < 0.05, t-test); (B) the decreased levels of CoQ10H2 were observed in both TCDD-treated mice (LL & HL groups), although no significant changes compared to control. (C) Altered CoQ10 redox state was pronouncedly elevated in mice exposed to TCDD compared to the control group (p < 0.05, t-test). (D) Proposed mitochondrial electron transport chain complex dysfunction in the mice exposed to TCDD (TCDD: 2,3,7,8-tetrachlorodibenzo-p-dioxin; SDH: succinate dehydrogenase; ETC: electron transport chain). Fig. 6
Fig. 5 One-step extraction of CoQ10H2 and CoQ10 with different alcohols (MeOH, methanol; EtOH, ethanol; and iPrOH, isopropanol). For CoQ10H2 extraction, MeOH was the most effective solvent (p < 0.05, t-test), meanwhile no significant difference was found for CoQ10 extraction.
This journal is © The Royal Society of Chemistry 2014
Analyst, 2014, 139, 5600–5604 | 5603
View Article Online
Published on 31 July 2014. Downloaded by Christian Albrechts Universitat zu Kiel on 28/10/2014 12:14:23.
Analyst
form of CoQ10) was 0.241 0.012 ng per 30 mg freeze dried tissue and 0.245 0.040 ng per 30 mg freeze dried tissue in lowand high-dose exposed groups, respectively, signicantly lower than that in the control group (0.347 0.073 ng per 30 mg freeze dried tissue, p < 0.005, t-test). Although levels of CoQ10H2 were not altered signicantly (Fig. 6B), the ratios of CoQ10H2 to total coenzyme Q10 were pronouncedly elevated (Fig. 6C). Together, these data suggest that alteration of oxidative phosphorylation in mitochondria occurred in liver tissues of mice exposed to TCDD. In our previous metabolomics study,29 succinate dehydrogenase (SDH) which is critical in electron transport chain (ETC), was postulated to be inhibited in mice livers due to TCDD exposure. The results of both forms of CoQ10 are in good agreement with previous indication from the metabolomic data. Therefore, the redox state of CoQ10 suggests that abnormalities of respiratory electron transport chain complexes in mice are caused by TCDD exposure (Fig. 6D).
4 Conclusions A UPLC-MS/MS method for the simultaneous measurement of both redox forms of the coenzyme Q10 in mammalian tissues was developed and validated for the purpose of investigating the role of these compounds in pathophysiological processes. The dual-use of methanol, namely, as a sample extraction solvent and as a non-aqueous reversed mobile phase for chromatographic separation, enables quick sample preparation, short separation run time, and sensitive and selective tandem mass spectrometric detection. The method was successfully applied in an investigation of TCDD-induced toxicity showing signicantly decreased levels of total CoQ10, and elevated ratios of the reduced form of CoQ10H2 to the total CoQ10 content in mice exposed to TCDD, revealing the disturbed ETC in liver tissues.
Acknowledgements This work was supported by Hong Kong Research Grant Council Collaborative Research Fund (HKBU5/CRF/10) and the National Natural Science Foundation of China (NSFC-21377106).
Notes and references 1 F. L. Crane, J. Am. Coll. Nutr., 2001, 20, 591–598. 2 D. A. Groneberg, B. Kindermann, M. Althammer, M. Klapper, J. Vormann, G. P. Littarru and F. D¨ oring, Int. J. Biochem. Cell Biol., 2005, 37, 1208–1218. 3 S. K. Lee, J. O. Lee, J. H. Kim, N. Kim, G. Y. You, J. W. Moon, J. Sha, S. J. Kim, Y. W. Lee, H. J. Kang, S. H. Park and H. S. Kim, Cell. Signalling, 2012, 24, 2329–2336. 4 R. V. Cooney, Q. Dai, Y. T. Gao, W. H. Chow, A. A. Franke, X. O. Shu, H. Li, B. Ji, Q. Cai, W. Chai and W. Zheng, Cancer Epidemiol., Biomarkers Prev., 2011, 20, 1124–1130. 5 M. Mezawa, M. Takemoto, S. Onishi, R. Ishibashi, T. Ishikawa, M. Yamaga, M. Fujimoto, E. Okabe, P. He, K. Kobayashi and K. Yokote, BioFactors, 2012, 38, 416–421. 6 Multiple-System Atrophy Research Collaboration, N. Engl. J. Med., 2013, 369, 233–244.
5604 | Analyst, 2014, 139, 5600–5604
Paper
7 L. Roffe, K. Schmidt and E. Ernst, J. Clin. Oncol., 2004, 22, 4418–4424. 8 D. Graham, N. N. Huynh, C. A. Hamilton, E. Beattie, R. A. Smith, H. M. Cochem´ e, M. P. Murphy and A. F. Dominiczak, Hypertension, 2009, 54, 322–328. 9 T. J. Shi, M. D. Zhang, H. Zeberg, J. Nilsson, J. Gr¨ unler, S. X. Liu, Q. Xiang, J. Persson, K. J. Fried, S. B. Catrina, M. Watanabe, P. Arhem, K. Brismar and T. G. H¨ okfelt, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 690–695. 10 V. Mugoni, R. Postel, V. Catanzaro, E. De Luca, E. Turco, G. Digilio, L. Silengo, M. P. Murphy, C. Medana, D. Y. Stainier, J. Bakkers and M. M. Santoro, Cell, 2013, 152, 504–518. 11 G. Hansen, P. Christensen, E. T¨ uchsen and T. Lund, Analyst, 2004, 129, 45–50. 12 K. Teshima and T. Kondo, Anal. Biochem., 2005, 338, 12–19. 13 J. Ruiz-Jim´ enez, F. Priego-Capote, J. M. Mata-Granados, J. M. Quesada and M. D. Luque de Castro, J. Chromatogr. A, 2007, 1175, 242–248. 14 K. E. Duberley, I. P. Hargreaves, K. A. Chaiwatanasirikul, S. J. Heales, J. M. Land, S. Rahman, K. Mills and S. Eaton, Rapid Commun. Mass Spectrom., 2013, 27, 924–930. 15 O. Itkonen, A. Suomalainen and U. Turpeinen, Clin Chem., 2013, 59, 1260–1267. 16 P. H. Tang, M. V. Miles, A. DeGrauw, A. Hershey and A. Pesce, Clin. Chem., 2001, 47, 256–265. 17 P. Niklowitz, F. D¨ oring, M. Paulussen and T. Menke, Anal. Biochem., 2013, 437, 88–94. 18 P. H. Tang and M. V. Miles, Methods Mol. Biol., 2012, 837, 149–168. 19 A. A. Franke, C. M. Morrison, J. L. Bakke, L. J. Custer, X. Li and R. V. Cooney, Free Radical Biol. Med., 2010, 48, 1610– 1617. 20 J. Lagendijk, J. B. Ubbink and W. J. Vermaak, J. Lipid Res., 1996, 37, 67–75. 21 M. J. Turkowicz and J. Karpi´ nska, BioFactors, 2013, 39, 176– 185. 22 S. H. Hahn, S. Kerfoot and V. Vasta, Methods Mol. Biol., 2012, 837, 169–179. 23 B. Frei, M. C. Kim and B. N. Ames, Proc. Natl. Acad. Sci. U. S. A., 1990, 87, 4879–4883. 24 P. O. Edlund, J. Chromatogr., 1988, 425, 87–97. 25 F. Mosca, D. Fattorini, S. Bompadre and G. P. Littarru, Anal. Biochem., 2002, 305, 49–54. 26 U.S. Food and Drug Administration, Center for Drug Evaluation and Research, Guidance for Industry: Bioanalytical Method Validation, 2001. http://www.fda.gov/downloads/Drugs/ GuidanceComplianceRegulatoryInformation/Guidances/ ucm070107.pdf. 27 D. R. Boverhof, L. D. Burgoon, C. Tashiro, B. Chittim, J. R. Harkema, D. B. Jump and T. R. Zacharewski, Toxicol. Sci., 2005, 85, 1048–1063. 28 D. R. Boverhof, L. D. Burgoon, C. Tashiro, B. Sharratt, B. Chittim, J. R. Harkema, D. L. Mendrick and T. R. Zacharewski, Toxicol. Sci., 2006, 94, 398–416. 29 S. H. Lin, Z. Yang, H. D. Liu and Z. W. Cai, Mol. BioSyst., 2011, 7, 1956–1965.
This journal is © The Royal Society of Chemistry 2014
Published on 31 July 2014. Downloaded by Christian Albrechts Universitat zu Kiel on 28/10/2014 12:14:23.
PAPER
View Journal | View Issue
Cite this: Analyst, 2014, 139, 5600
Rapid assessment of the coenzyme Q10 redox state using ultrahigh performance liquid chromatography tandem mass spectrometry† Zhi Tang,a Shangfu Li,a Xinyuan Guan,b Philippe Schmitt-Kopplin,c Shuhai Lin*a and Zongwei Cai*a An improved method for accurate and rapid assessment of the coenzyme Q10 (CoQ10) redox state using ultrahigh performance liquid chromatography tandem mass spectrometry was described, with particular attention given to the instability of the reduced form of CoQ10 during sample preparation, chromatographic separation and mass spectrometric detection. As highly lipophilic compounds in complex biological matrices, both reduced and oxidized forms of CoQ10 were extracted simultaneously from the tissue samples by methanol which is superior to ethanol and isopropanol. After centrifugation, the supernatants were immediately separated on a C18 column with isocratic elution using methanol containing 2 mM ammonium acetate as a non-aqueous mobile phase, and detected by positive electrospray ionization tandem mass spectrometry in multiple reaction monitoring (MRM) mode. Ammonium acetate as an additive in methanol provided enhanced mass spectrometric responses for both forms of CoQ10, primarily due to stable formation of adduct ions [M + NH4]+, which served as precursor ions in positive ionization MRM transitions. The assay showed a linear range of 8.6–8585 ng mL1 for CoQ10H2 and 8.6–4292 ng mL1 for CoQ10. The limits of detection (LODs) were 7.0 and 1.0 ng mL1 and limits of quantification (LOQs) were 15.0 and 5.0 ng mL1 for CoQ10H2 and CoQ10, respectively. This rapid
Received 29th April 2014 Accepted 31st July 2014
extractive and analytical method could avoid artificial auto-oxidation of the reduced form of CoQ10, enabling the native redox state assessment. This reliable method was also successfully applied for the
DOI: 10.1039/c4an00760c
measurement of the CoQ10 redox state in liver tissues of mice exposed to 2,3,7,8-tetrachlorodibenzo-p-
www.rsc.org/analyst
dioxin, revealing the down-regulated mitochondrial electron transport chain.
1
Introduction
Endogenous coenzyme Q10 (CoQ10) is a highly lipophilic small molecule consisting of a benzoquinone head group and 10 isoprenyl subunits in its side chain, which coexists in two different types – the reduced form (CoQ10H2, ubiquinol) and the oxidized form (CoQ10, ubiquinone) (Fig. 1), which is present mainly in the mitochondria of eukaryotic cells. As a crucial component in the mitochondrial respiratory chain for electron transport and cellular energy production,1 CoQ10 plays an important role in a wide range of biochemical, physiological and pathological processes. For example, it affects gene expression involved in cell metabolism,2,3 and reduced levels of
CoQ10 are implicated in different diseases, like breast cancer,4 diabetes,5 and neurodegenerative diseases.6 Given the vital role of CoQ10 in biological functions, its use as a nutritional supplement or adjuvant therapy for various diseases, including cancer,7 cardiovascular disease,8 and diabetes9 and its possible effect of limiting the side effects of statin therapy,10 have also been reported in recent investigations. It has become obvious that the unique antioxidant and health-promoting properties of CoQ10 are largely dependent both on its concentration and
a
State Key Laboratory of Environmental and Biological Analysis, Department of Chemistry, Hong Kong Baptist University, Hong Kong SAR, China. E-mail: zwcai@ hkbu.edu.hk; [email protected]; Fax: +852 3411 7348; Tel: +852 3411 7070
b
Department of Clinical Oncology, University of Hong Kong, Hong Kong SAR, China
c
Analytical Biogeochemistry, Helmholtz-Zentrum Muenchen–German Research Center for Environmental Health, Ingolstaedter Landstrasse 1, D-85764, Oberschleißheim, Germany † Electronic supplementary 10.1039/c4an00760c
information
5600 | Analyst, 2014, 139, 5600–5604
(ESI)
available.
See
DOI:
Fig. 1 Reduced form of CoQ10 can be easily oxidized to its corresponding oxidized form (left: reduced form, CoQ10H2, ubiquinol; right: oxidized form, CoQ10, ubiquinone).
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 31 July 2014. Downloaded by Christian Albrechts Universitat zu Kiel on 28/10/2014 12:14:23.
Paper
redox state. To advance the fundamental understanding of the roles of CoQ10 in various physiological and pathological conditions, as well as an efficiency evaluation of its supplementation, an analytical method for assessing the CoQ10 content and its redox state in complex biological matrices is required. Therefore, the aim of the current study was to develop a specic and sensitive mass spectrometry-based method for rapid quantication of CoQ10 in biological samples. In previous studies, CoQ10 concentrations in different biological specimens, including serum, cells, subcellular organelles, or other clinical and research samples were analyzed by high performance liquid chromatography (HPLC) coupled with a tandem mass spectrometer11–15 or an electrochemical detector.16–18 Unlike HPLC coupled with electrochemical detection, ultrahigh performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) would be a newly developed method with high sensitivity and high selectivity to measure both forms of CoQ10. However, due to the instability of the reduced form of CoQ10 which can be easily oxidized to its corresponding oxidized form during sample pre-treatment,19–21 reliable assay of the coenzyme should be further explored. With particular attention to this problem, here, we described a rapid quantitative assay for characterization of the CoQ10 redox state in mouse liver tissue samples using UPLC-MS/MS.
2 Experimental 2.1
Chemicals and reagents
2,3,7,8-Tetrachlorodibenzo-p-dioxin (abbr. TCDD, purity >99%) was obtained from AccuStandard, Inc. (New Haven, CT, USA). Coenzyme Q10 standard (oxidized form, >98%, HPLC), and sodium borohydride (NaBH4) were obtained from SigmaAldrich (St. Louis, MO, USA) and used without further purication. Dipropoxy-CoQ10 (DP-Q10), a propyl analogue of CoQ10 served as the internal standard, was prepared according to a recently reported protocol.22 Methanol, ethanol, isopropanol, and hexane were of HPLC grade and were obtained from Tedia (Faireld, OH, USA). Ultrapure water was puried by using a Milli-Q Academic Water Purication System from Millipore (Millipore, Bedford, MA, USA). 2.2
Instrumentation and UPLC-MS/MS analysis
LC-MS/MS analyses were performed on a Waters ACQUITY ultrahigh performance liquid chromatograph (UPLC) system hyphenated with a TQD triple quadrupole mass spectrometer tted with a Z-spray electrospray ionization source. The chromatographic separation was achieved by using a C18 reverse phase column (ACQUITY UPLC BEH C18 column, 2.1 mm 50 mm, 1.7 mm) employing methanol containing 2 mM ammonium acetate as the mobile phase at the ow rate of 0.4 mL min1 under isocratic elution conditions. The column temperature was kept at 40 C. The mass spectrometer was operated in positive ionization mode. The optimized parameters were capillary voltage, 2 kV; cone voltage, 30 V; source and desolvation temperatures were 120 C and 400 C, respectively. Nitrogen gas was used as the desolvation gas (700 L h1) and
This journal is © The Royal Society of Chemistry 2014
Analyst
cone gas (50 L h1). Argon gas was used for collision-induced dissociation. Multiple reaction monitoring (MRM) was used for quantication of both reduced and oxidized CoQ10. MRM transitions, cone voltages and collision energies as well as dwell times are summarized in Table 1. MassLynx v.4.1 soware was used for data acquisition and analysis (Waters Corp., Milford, MA, USA). 2.3
Animal experiment
Eight week old male C57BL/6J mice were purchased from the Chinese University of Hong Kong. The animal experimental procedures were approved by the Committee on the Use of Live Animals for Teaching and Research, Department of Health, Hong Kong SAR, China and were carried out in accordance with the NIH Guide for the Care and Use of Laboratory Animals. The mice were divided into control groups (n ¼ 7) and TCDD treated groups (n ¼ 7 for the high dosage group; n ¼ 7 for the low dosage group). Mice in each group were treated with either corn oil (the control group) or TCDD dissolved in corn oil (2.0 mg mL1 and 0.2 mg mL1) for three consecutive days by oral administration of 20 mg kg1 and 2.0 mg kg1 body weight, respectively. On the tenth day, all the mice were decapitated, and the liver tissues were harvested, snap frozen with liquid nitrogen and then stored at 80 C. 2.4
Standard solution preparation
One milligram per milliliter CoQ10 stock solution was prepared as follows: 25 mg of CoQ10 standard was accurately weighed and placed into a 25 mL volumetric ask. Three milliliters hexane as the co-solvent was added to dissolve the solid completely, and the resultant solution was diluted to the mark with isopropanol and stored at 20 C. Working solutions of CoQ10 were prepared by diluting the stock solution with nitrogen-purged methanol. As the reference standard of CoQ10H2 (the reduced form of CoQ10) was not commercially available, its solutions were freshly prepared from the corresponding oxidized form upon reduction with sodium borohydride, as described by Frei et al.23 Briey, 200 mL of the CoQ10 stock solution was placed into an Eppendorf tube and 1.0 mg of solid NaBH4 was added. The mixture was then vortexed at room temperature. CoQ10 was converted quantitatively to CoQ10H2, as indicated by disappearance of ubiquinone and the appearance of product monitored by UPLC-MS/MS. Because of the instability of CoQ10H2, its solutions were prepared immediately prior to the analysis, and freshly prepared each time, and the working standard solutions
MRM transitions used for quantification of the reduced form and oxidized form of coenzyme Q10 and internal standard DP-Q10
Table 1
Analytes
MRM transition (m/z)
Dwell time (s)
CV (V)
CE (eV)
CoQ10H2 CoQ10 DP-Q10
883.0 / 197.0 881.0 / 197.0 937.0 / 253.0
0.1 0.1 0.1
30 30 30
25 25 20
Analyst, 2014, 139, 5600–5604 | 5601
View Article Online
Analyst
Paper
were generated by subsequent serial dilutions using deoxygenated methanol.
Published on 31 July 2014. Downloaded by Christian Albrechts Universitat zu Kiel on 28/10/2014 12:14:23.
2.5
Tissue sample pre-treatment and CoQ10 extraction
To avoid articial oxidation of the analytes, all solvents were deoxygenated shortly before use by purging with dry nitrogen and kept on ice. Three different alcoholic solvents, including methanol, ethanol, and isopropanol were tested to obtain optimal extraction efficiency. Each tissue was freeze-dried, pulverized, and thoroughly mixed, and then 30 mg freeze-dried powders were extracted with 300 mL ice-chilled alcoholic solvent containing 30 ng internal standard. The mixture was vortexed for 30 seconds and placed on ice for 15 min and then centrifuged at 12 000 rpm at 4 C for 10 min. This process was repeated twice; the supernatants were pooled together, then transferred into an auto-sampler vial and analyzed immediately. The remaining sample was stored at 20 C for a reproducibility study. 2.6
Method validation
Fig. 2 The full scan mass spectrum of oxidized CoQ10 (A) and product ion mass spectra of [M + NH4]+ (B).
Method validation was carried out in terms of precision, linearity, limits of detection, limits of quantitation, matrix effects, recovery and stability. Calibration curves were built using serial dilutions of a 1.0 mg mL1 CoQ10 standard solution, spiked with in-house made internal standard DP-Q10. The concentration levels of CoQ10 spanned from below 10 ng mL1 to near 10 000 ng mL1 while the concentration of the internal standard was maintained constant at 100 ng mL1. Three replicate injections were made for each calibration level.
aqueous reversed-phase mobile phase for chromatographic separation of CoQ10 is not only readily compatible with electrospray ionization mass spectrometry but also with tissue sample preparation. To achieve a very short run time in UPLC analysis as well as reproducible results, the column temperature was controlled, and temperatures ranging from 25 C to 40 C at an interval of ve degree were tested. As shorter retention times at higher column temperature were obtained, the column was maintained at 40 C during the following experiments (Fig. 4).
3 Results and discussion
3.2
3.1
UPLC-ESI-MS/MS method development
An improved UPLC-ESI-MS/MS method for accurate and rapid assessment of CoQ10 in mouse liver tissue samples was developed and validated, with particular attention given to chemical instability of CoQ10H2, and the low biological concentration levels present in complex matrixes. 3.1.1 ESI-MS/MS optimization. The MS/MS parameters were optimized by direct injection using the standard compounds. The most intense fragment ion in the product ion spectrum for each standard was selected to be monitored in MRM mode. The cone voltage (CV) and collision energy (CE) were optimized for MRM transitions of each [M + NH4]+ ion, which were predominantly observed in the positive ESI spectra (Fig. 2). The MRM transitions, cone voltages, collision energies, and internal standards are listed in Table 1. 3.1.2 UPLC separation optimization. The UPLC method was optimized to ensure resolution between the two types of coenzyme Q10. Different organic solvents, including methanol and acetonitrile as well as their combination were tested, and it was found that using methanol as the mobile phase provided well-resolved peaks with good peak shapes over the acetonitrile– methanol (1 : 4, v/v) solvent system (Fig. 3). The addition of 2 mM ammonium acetate as a mobile phase modier provided favorable sensitivities as [M + NH4]+ ions were used as parent ions in MRM mode. To be duly noted, methanol as a non-
5602 | Analyst, 2014, 139, 5600–5604
Coenzyme Q10 extraction optimization
Sample preparation is one of the major factors affecting the quantitative estimation of CoQ10 in biological samples. As reported previously,24,25 the use of alcohol for CoQ10 extraction combined with in situ chemical oxidation and followed by direct HPLC analysis is a fast and robust method for evaluating total CoQ10 in plasma. In this work, the efficiency of different alcohols, including methanol, ethanol, and isopropanol for the extraction of both forms of CoQ10, was evaluated. Methanol was found to be the most efficient extraction solvent for both forms
Fig. 3 Mobile phase composition on separation of the two forms of CoQ10 (A) methanol as the mobile phase results in well-resolved peaks; (B) acetonitrile–methanol (1 : 4, v/v) provides no separation.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 31 July 2014. Downloaded by Christian Albrechts Universitat zu Kiel on 28/10/2014 12:14:23.
Paper
Analyst
both forms of CoQ10 was achieved as indicated by imprecision less than 10% relative standard deviation. The relatively lower but consistent recovery (79%) for the spiked CoQ10 sample at the concentration of 45 ng mL1 (Table S2†) could be in part due to adsorption losses during sample preparation (for example, binding to biological matrices or being adsorbed on sample containers such as tubes or injection vials), but it is still practically acceptable. Calibration curves for CoQ10H2 and CoQ10 were linear over the range of 8.6 to 8585 ng mL1 and 8.6 to 4292 ng mL1 (Fig. S1 & S2†). The limits of detection (LODs) were 7.0 and 1.0, and the limits of quantitation (LOQs) were 15.0 and 5.0 ng mL1 for CoQ10H2 and CoQ10, respectively (Table S2†). More detailed results of the method performance and validation are provided in the ESI (Table S1, Fig. S3 and S4†).
3.4
Fig. 4 Column temperature on retention of both forms of CoQ10 and the internal standard (left to right: the 1st peak, CoQ10H2; the 2nd one, CoQ10; the last one, the internal standard DP-Q10).
of CoQ10 in our testing (Fig. 5). Thus, the optimized sample preparation method is single-step extraction of CoQ10 using methanol, which is compatible with chromatographic separation and mass spectrometric detection, eliminating liquid– liquid extraction and further concentration steps.
3.3
Method validation
Biological application
Previous studies have revealed that TCDD damaged the mouse liver in a dose-dependent manner.27,28 Abnormal levels of serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) activities were reported in TCDD-exposed mice. However, the underlying mechanisms of TCDD-induced hepatotoxicity in mice and humans are still not fully understood. The redox state in the liver may also reveal the toxic effect of TCDD exposure in mice. To measure whether the CoQ10 redox state was altered in mice exposed to TCDD, we applied the validated method to liver tissue samples, obtained from three different experimental mouse groups that were treated with low-dosage (LL), highdosage TCDD (HL), and control (CL). As shown in Fig. 6A, the amount of total CoQ10 (sum of the reduced form and oxidized
The developed method was validated in terms of precision, sensitivity, matrix effects, recovery, linearity, limit of detection (LOD), and limit of quantitation (LOQ) according to general guidelines for validation of the bioanalytical method from US Food and Drug Administration (FDA).26 Good reproducibility (precision at n ¼ 9, three parallel samples at low, medium, and high concentration levels, and each measured three times) of
Altered CoQ10 redox state with significant changes in liver tissues of mice exposed to TCDD. (A) Significantly decreased levels of the total CoQ10 in mice exposed to TCDD (CL vs. LL, and CL vs. HL, p < 0.05, t-test); (B) the decreased levels of CoQ10H2 were observed in both TCDD-treated mice (LL & HL groups), although no significant changes compared to control. (C) Altered CoQ10 redox state was pronouncedly elevated in mice exposed to TCDD compared to the control group (p < 0.05, t-test). (D) Proposed mitochondrial electron transport chain complex dysfunction in the mice exposed to TCDD (TCDD: 2,3,7,8-tetrachlorodibenzo-p-dioxin; SDH: succinate dehydrogenase; ETC: electron transport chain). Fig. 6
Fig. 5 One-step extraction of CoQ10H2 and CoQ10 with different alcohols (MeOH, methanol; EtOH, ethanol; and iPrOH, isopropanol). For CoQ10H2 extraction, MeOH was the most effective solvent (p < 0.05, t-test), meanwhile no significant difference was found for CoQ10 extraction.
This journal is © The Royal Society of Chemistry 2014
Analyst, 2014, 139, 5600–5604 | 5603
View Article Online
Published on 31 July 2014. Downloaded by Christian Albrechts Universitat zu Kiel on 28/10/2014 12:14:23.
Analyst
form of CoQ10) was 0.241 0.012 ng per 30 mg freeze dried tissue and 0.245 0.040 ng per 30 mg freeze dried tissue in lowand high-dose exposed groups, respectively, signicantly lower than that in the control group (0.347 0.073 ng per 30 mg freeze dried tissue, p < 0.005, t-test). Although levels of CoQ10H2 were not altered signicantly (Fig. 6B), the ratios of CoQ10H2 to total coenzyme Q10 were pronouncedly elevated (Fig. 6C). Together, these data suggest that alteration of oxidative phosphorylation in mitochondria occurred in liver tissues of mice exposed to TCDD. In our previous metabolomics study,29 succinate dehydrogenase (SDH) which is critical in electron transport chain (ETC), was postulated to be inhibited in mice livers due to TCDD exposure. The results of both forms of CoQ10 are in good agreement with previous indication from the metabolomic data. Therefore, the redox state of CoQ10 suggests that abnormalities of respiratory electron transport chain complexes in mice are caused by TCDD exposure (Fig. 6D).
4 Conclusions A UPLC-MS/MS method for the simultaneous measurement of both redox forms of the coenzyme Q10 in mammalian tissues was developed and validated for the purpose of investigating the role of these compounds in pathophysiological processes. The dual-use of methanol, namely, as a sample extraction solvent and as a non-aqueous reversed mobile phase for chromatographic separation, enables quick sample preparation, short separation run time, and sensitive and selective tandem mass spectrometric detection. The method was successfully applied in an investigation of TCDD-induced toxicity showing signicantly decreased levels of total CoQ10, and elevated ratios of the reduced form of CoQ10H2 to the total CoQ10 content in mice exposed to TCDD, revealing the disturbed ETC in liver tissues.
Acknowledgements This work was supported by Hong Kong Research Grant Council Collaborative Research Fund (HKBU5/CRF/10) and the National Natural Science Foundation of China (NSFC-21377106).
Notes and references 1 F. L. Crane, J. Am. Coll. Nutr., 2001, 20, 591–598. 2 D. A. Groneberg, B. Kindermann, M. Althammer, M. Klapper, J. Vormann, G. P. Littarru and F. D¨ oring, Int. J. Biochem. Cell Biol., 2005, 37, 1208–1218. 3 S. K. Lee, J. O. Lee, J. H. Kim, N. Kim, G. Y. You, J. W. Moon, J. Sha, S. J. Kim, Y. W. Lee, H. J. Kang, S. H. Park and H. S. Kim, Cell. Signalling, 2012, 24, 2329–2336. 4 R. V. Cooney, Q. Dai, Y. T. Gao, W. H. Chow, A. A. Franke, X. O. Shu, H. Li, B. Ji, Q. Cai, W. Chai and W. Zheng, Cancer Epidemiol., Biomarkers Prev., 2011, 20, 1124–1130. 5 M. Mezawa, M. Takemoto, S. Onishi, R. Ishibashi, T. Ishikawa, M. Yamaga, M. Fujimoto, E. Okabe, P. He, K. Kobayashi and K. Yokote, BioFactors, 2012, 38, 416–421. 6 Multiple-System Atrophy Research Collaboration, N. Engl. J. Med., 2013, 369, 233–244.
5604 | Analyst, 2014, 139, 5600–5604
Paper
7 L. Roffe, K. Schmidt and E. Ernst, J. Clin. Oncol., 2004, 22, 4418–4424. 8 D. Graham, N. N. Huynh, C. A. Hamilton, E. Beattie, R. A. Smith, H. M. Cochem´ e, M. P. Murphy and A. F. Dominiczak, Hypertension, 2009, 54, 322–328. 9 T. J. Shi, M. D. Zhang, H. Zeberg, J. Nilsson, J. Gr¨ unler, S. X. Liu, Q. Xiang, J. Persson, K. J. Fried, S. B. Catrina, M. Watanabe, P. Arhem, K. Brismar and T. G. H¨ okfelt, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 690–695. 10 V. Mugoni, R. Postel, V. Catanzaro, E. De Luca, E. Turco, G. Digilio, L. Silengo, M. P. Murphy, C. Medana, D. Y. Stainier, J. Bakkers and M. M. Santoro, Cell, 2013, 152, 504–518. 11 G. Hansen, P. Christensen, E. T¨ uchsen and T. Lund, Analyst, 2004, 129, 45–50. 12 K. Teshima and T. Kondo, Anal. Biochem., 2005, 338, 12–19. 13 J. Ruiz-Jim´ enez, F. Priego-Capote, J. M. Mata-Granados, J. M. Quesada and M. D. Luque de Castro, J. Chromatogr. A, 2007, 1175, 242–248. 14 K. E. Duberley, I. P. Hargreaves, K. A. Chaiwatanasirikul, S. J. Heales, J. M. Land, S. Rahman, K. Mills and S. Eaton, Rapid Commun. Mass Spectrom., 2013, 27, 924–930. 15 O. Itkonen, A. Suomalainen and U. Turpeinen, Clin Chem., 2013, 59, 1260–1267. 16 P. H. Tang, M. V. Miles, A. DeGrauw, A. Hershey and A. Pesce, Clin. Chem., 2001, 47, 256–265. 17 P. Niklowitz, F. D¨ oring, M. Paulussen and T. Menke, Anal. Biochem., 2013, 437, 88–94. 18 P. H. Tang and M. V. Miles, Methods Mol. Biol., 2012, 837, 149–168. 19 A. A. Franke, C. M. Morrison, J. L. Bakke, L. J. Custer, X. Li and R. V. Cooney, Free Radical Biol. Med., 2010, 48, 1610– 1617. 20 J. Lagendijk, J. B. Ubbink and W. J. Vermaak, J. Lipid Res., 1996, 37, 67–75. 21 M. J. Turkowicz and J. Karpi´ nska, BioFactors, 2013, 39, 176– 185. 22 S. H. Hahn, S. Kerfoot and V. Vasta, Methods Mol. Biol., 2012, 837, 169–179. 23 B. Frei, M. C. Kim and B. N. Ames, Proc. Natl. Acad. Sci. U. S. A., 1990, 87, 4879–4883. 24 P. O. Edlund, J. Chromatogr., 1988, 425, 87–97. 25 F. Mosca, D. Fattorini, S. Bompadre and G. P. Littarru, Anal. Biochem., 2002, 305, 49–54. 26 U.S. Food and Drug Administration, Center for Drug Evaluation and Research, Guidance for Industry: Bioanalytical Method Validation, 2001. http://www.fda.gov/downloads/Drugs/ GuidanceComplianceRegulatoryInformation/Guidances/ ucm070107.pdf. 27 D. R. Boverhof, L. D. Burgoon, C. Tashiro, B. Chittim, J. R. Harkema, D. B. Jump and T. R. Zacharewski, Toxicol. Sci., 2005, 85, 1048–1063. 28 D. R. Boverhof, L. D. Burgoon, C. Tashiro, B. Sharratt, B. Chittim, J. R. Harkema, D. L. Mendrick and T. R. Zacharewski, Toxicol. Sci., 2006, 94, 398–416. 29 S. H. Lin, Z. Yang, H. D. Liu and Z. W. Cai, Mol. BioSyst., 2011, 7, 1956–1965.
This journal is © The Royal Society of Chemistry 2014
