Anal. Chem. 2010, 82, 5282–5289
Ultra Performance Liquid Chromatography-Mass Spectrometry Profiling of Bile Acid Metabolites in Biofluids: Application to Experimental Toxicology Studies Elizabeth J. Want,*,† Muireann Coen,† Perrine Masson,† Hector C. Keun,† Jake T. M. Pearce,† Michael D. Reily,‡ Donald G. Robertson,‡ Cynthia M. Rohde,§ Elaine Holmes,† John C. Lindon,† Robert S. Plumb,| and Jeremy K. Nicholson† Biomolecular Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, SW7 2AZ, U.K., Bristol Myers-Squibb, Route 206 and Province Line Road, Princeton, New Jersey 08543-4000, Drug Safety Research and Development, Pfizer Global Research and Development, Chazy, New York 12921, and Waters Corporation, 34 Maple Street, Milford, Massachusetts 01757 We have developed an ultra performance liquid chromatography-mass spectrometry (UPLC-MSE) method to measure bile acids (BAs) reproducibly and reliably in biological fluids and have applied this approach for indications of hepatic damage in experimental toxicity studies. BAs were extracted from serum using methanol, and an Acquity HSS column coupled to a Q-ToF mass spectrometer was used to separate and identify 25 individual BAs within 5 min. Employing a gradient elution of water and acetonitrile over 21 min enabled the detection of a wide range of endogenous metabolites, including the BAs. The utilization of MSE allowed for characteristic fragmentation information to be obtained in a single analytical run, easily distinguishing glycine and taurine BA conjugates. The proportions of these conjugates were altered markedly in an experimental toxic state induced by galactosamine exposure in rats. Principally, taurine-conjugated BAs were greatly elevated (∼50-fold from control levels), and were highly correlated to liver damage severity as assessed by histopathological scoring (r ) 0.83), indicating their potential as a sensitive measure of hepatic damage. The UPLC-MS approach to BA analysis offers a sensitive and reproducible tool that will be of great value in exploring both markers and mechanisms of hepatotoxicity and can readily be extended to clinical studies of liver damage. Hepatic dysfunction is known to be associated with disturbances in bile acid metabolism.1-9 Formed from cholesterol in the liver, bile acids (BAs) are excreted in mammalian bile mainly * Corresponding author: (e-mail)
[email protected]. † Imperial College London. ‡ Bristol Myers-Squibb. § Pfizer Global Research and Development. | Waters Corp. (1) Vlahcevic, Z. R.; Prugh, M. F.; Gregory, D. H.; Swell, L. Clin. Gastroenterol. 1977, 6 (1), 25–43. (2) Stiehl, A. Clin. Gastroenterol. 1977, 6 (1), 45–67. (3) Hofmann, A. F. Arch. Intern. Med. 1999, 159 (22), 2647–2658.
5282
Analytical Chemistry, Vol. 82, No. 12, June 15, 2010
as glycine or taurine conjugates. BAs then undergo further metabolism by bacterial and hepatic enzymes during enterohepatic circulation, and it is well-known that BA metabolism is codependent on the biological activities of the gut microflora.10-13 In normal human individuals, BAs are detected predominantly in bile and feces, with considerably lower urine or serum levels. However, in hepatobiliary and intestinal diseases, or following hepatic injury, disturbed BA enterohepatic circulation results in quantitative and qualitative serum BA changes,14-16 disrupting cholesterol synthesis and metabolism, thereby affecting BA concentrations and (4) Akashi, Y.; Miyazaki, H.; Yanagisawa, J.; Nakayama, F. Clin. Chim. Acta 1987, 168 (2), 199–206. (5) Bijleveld, C. M.; Vonk, R. J.; Kuipers, F.; Havinga, R.; Boverhof, R.; Koopman, B. J.; Wolthers, B. G.; Fernandes, J. Gastroenterology 1989, 97 (2), 427–432. (6) Shoda, J.; Tanaka, N.; Osuga, T.; Matsuura, K.; Miyazaki, H. J. Lipid Res. 1990, 31 (2), 249–259. (7) Erickson, R. P.; Bhattacharyya, A.; Hunter, R. J.; Heidenreich, R. A.; Cherrington, N. J. Am. J. Physiol. Gastrointest. Liver Physiol. 2005, 289 (2), G300-7. (8) Yousef, I. M. Bouchard, G.; Tuchweber, B.; Plaa, G. L. In Toxicology of the Liver, 2nd ed.; Plaa, G. L., Hewitt, W. R. , Eds.; Taylor & Francis: New York, 1998; pp347-382. (9) Ostrow, J. D. In Hepatic Transport and Bile Secretion: Physiology and Pathophysiology; Tavolini, N., Berk, P. D., Eds.; Raven Press: New York, 1993; pp 673-712. (10) Martin, F. P.; Dumas, M. E.; Wang, Y.; Legido-Quigley, C.; Yap, I. K.; Tang, H.; Zirah, S.; Murphy, G. M.; Cloarec, O.; Lindon, J. C.; Sprenger, N.; Fay, L. B.; Kochhar, S.; van Bladeren, P.; Holmes, E.; Nicholson, J. K. Mol. Syst. Biol. 2007, 3, 112. (11) Martin, F. P.; Wang, Y.; Sprenger, N.; Yap, I. K.; Rezzi, S.; Ramadan, Z.; Pere´-Trepat, E.; Rochat, F.; Cherbut, C.; van Bladeren, P.; Fay, L. B.; Kochhar, S.; Lindon, J. C.; Holmes, E.; Nicholson, J. K. Mol. Syst. Biol. 2008, 4, 205. (12) Martin, F. P.; Wang, Y.; Sprenger, N.; Yap, I. K.; Lundstedt, T.; Lek, P.; Rezzi, S.; Ramadan, Z.; van Bladeren, P.; Fay, L. B.; Kochhar, S.; Lindon, J. C.; Holmes, E.; Nicholson, J. K. Mol. Syst. Biol. 2008, 4, 157. (13) Claus, S. P.; Tsang, T. M.; Wang, Y.; Cloarec, O.; Skordi, E.; Martin, F. P.; Rezzi, S.; Ross, A.; Kochhar, S.; Holmes, E.; Nicholson, J. K. Mol Syst Biol. 2008, 4, 219. (14) Anwer, M. S.; Meyer, D. J. Vet. Clin. North Am. Small Anim. Pract. 1995, 25 (2), 503–517. (15) Ambros-Rudolph, C. M.; Glatz, M.; Trauner, M.; Kerl, H.; Mu ¨ llegger, R. R. Arch. Dermatol. 2007, 143 (6), 757–762. (16) Crosignani, A.; Del Puppo, M.; Longo, M.; De Fabiani, E.; Caruso, D.; Zuin, M.; Podda, M.; Javitt, N. B.; Kienle, M. G. Clin. Chim. Acta 2007, 382 (1-2), 82–88. 10.1021/ac1007078 2010 American Chemical Society Published on Web 05/14/2010
profiles in the liver, serum, urine, feces, and gallbladder.8,9 Serum BAs have been shown to be sensitive markers of several liver diseases, including chronic viral hepatitis, and intestinal disease in humans.2,17,18 Further, BAs can be markedly elevated in the urine and serum of rats treated with hepatotoxins, e.g., R-naphthyl isothiocyanate (ANIT) and carbon tetrachloride.19-24 Metabonomics can provide global insight into the actions of hepatotoxins on a whole organism, offering an untargeted, unbiased view of processes occurring at the system level.25,26 Several studies have applied 1H NMR spectroscopy to the interrogation of the hepatotoxins ANIT, D-(+)-galactosamine (galN), and butylated hydroxytoluene (BHT), and subsequent mathematical modeling indicated that an altered BA profile was a key characteristic of the toxic response.22,24,27 Other metabonomic studies have probed the actions of allyl alcohol,28 bromobenzene,29,30 and carbon tetrachloride,31 providing fresh metabolic insights. Although broad metabolic screening strategies have the potential to uncover biomarkers without a priori selection of an analyte set, where a particular class of molecular species is known to be diagnostic for a condition, targeted assays are often more sensitive and/or efficient. Traditional methods of measuring BAs have focused on total serum concentrations,32,33 but the ratio of taurine to glycine conjugation, as well as of the different classes of BAs (e.g., mono-, di-, and trihydroxylated) and the exact species of BA can yield important biological and clinically relevant information.8,9,34 BAs undergo hydrolysis, oxidation and reduction, isomerization, dehydroxylation, and hydroxylation reactions,35 in addition (17) Shima, T.; Tada, H.; Morimoto, M.; Nakagawa, Y.; Obata, H.; Sasaki, T.; Park, H.; Nakajo, S.; Nakashima, T.; Okanoue, T.; Kashima, K. Gastroenterol. Hepatol. 2000, 15 (3), 294–299. (18) de Caestecker, J. S.; Jazrawi, R. P.; Nisbett, J. A.; Joseph, A. E.; Maxwell, J. D.; Northfield, T. C. Eur. J. Gastroenterol. Hepatol. 1995, 7 (10), 955– 961. (19) Wang, G. F.; Stacey, N. H. Toxicol. Appl. Pharmacol. 1990, 105 (2), 209– 215. (20) Neghab, M.; Stacey, N. H. Chem. Biol. Interact. 1996, 99 (1-3), 179–192. (21) Thompson, M. B.; Davis, D. G.; Morris, R. W. J. Lipid Res. 1993, 34 (4), 553–561. (22) Davis, D. G.; Thompson, M. B. J. Lipid Res. 1993, 34 (4), 651–661. (23) Bai, C.; Canfield, P. J.; Stacey, N. H. Toxicology 1992, 75 (3), 221–234. (24) Beckwith-Hall, B. M.; Nicholson, J. K.; Nicholls, A. W.; Foxall, P. J.; Lindon, J. C.; Connor, S. C.; Abdi, M.; Connelly, J.; Holmes, E. Chem. Res. Toxicol. 1998, 11, 260–272. (25) Nicholson, J. K.; Lindon, J. C.; Holmes, E. Xenobiotica 1999, 29, 1181– 1189. (26) Nicholson, J. K.; Wilson, I. D. Nat. Rev. Drug Discov. 2003, 2, 668–676. (27) Beckwith-Hall, B. M.; Thompson, N. A.; Nicholson, J. K.; Lindon, J. C.; Holmes, E. Analyst 2003, 128, 814–818. (28) Holmes, E.; Nicholls, A. W.; Lindon, J. C.; Ramos, S.; Spraul, M.; Neidig, P.; Connor, S. C.; Connelly, J.; Damment, S. J. P.; Haselden, J.; Nicholson, J. K. NMR Biomed. 1998, 11, 235–244. (29) Heijne, W. H. M.; Lamers, R-J.A.N.; Van Bladeren, P. J.; Groten, J. P.; Van Nesselrooij, J. H. J.; Van Ommen, B. Toxicol. Pathol. 2005, 33, 425–433. (30) Waters, N. J.; Waterfield, C. J.; Farrant, R. D.; Holmes, E.; Nicholson, J. K. J. Proteome Res. 2006, 5, 1448–1459. (31) Yanping, L.; Duanyun, S.; Zongpeng, Z.; Changxiao, L. Toxicology 2009, 256, 191–200. (32) Scalia, S. J. Chromatogr., B: Biomed. Sci. Appl. 1995, 671 (1-2), 299–317. (33) Roda, A.; Piazza, F.; Baraldini, M. J. Chromatogr., B: Biomed. Sci. Appl. 1998, 717 (1-2), 263–78. (34) Perwaiz, S.; Tuchweber, B.; Mignault, D.; Gilat, T.; Yousef, I. M. J. Lipid Res. 2001, 42 (1), 114–119. (35) Hofmann, A. F. In TheLiver: Biology and Pathobiology, 3rd ed.; Arias, I. M., Boyer, J. L., Fausto, N., Jakoby, W. B., Schachter, D. A., Shafritz, D. A., Eds.; Raven Press: New York; 1994; pp 677-718.
to phase 2 conjugation of the hydroxyl groups with glucuronic acid, sulfuric acid, glucose, or N-acetylglucosamine36,37 in the liver, intestine, and kidney. This complexity of metabolism, the typically low concentration (µmolar) of BAs in biological fluids, and the existence of multiple isobaric structural isomers make BA separation and quantitation challenging.9 Additionally, broad differences in physicochemical properties, e.g., lipophilicity and polarity, can exist within each BA class. Hence, sophisticated chromatographic techniques are required in order to obtain comprehensive BA profiles in disordered states. Gas chromatography-mass spectrometry provides a sensitive, specific and reproducible method for BA detection and quantitation, but disadvantages include lengthy derivatization steps and long analysis times.38 Liquid chromatography (LC) methods can be used with ultraviolet (UV)39 or fluorescence detection, but disadvantages include (a) unreliability of UV with regard to selectivitysBAs lack a strong chromophoresand (b) the need for separation of BAs by conjugation class, as well as hydrolysis and derivatization prior to fluorescence detection. LC coupled to mass spectrometry (MS) offers a more sensitive approach for BA analysis.40-44 Ultra performance liquid chromatography (UPLC),45 which employs sub-2-µm particle columns,46 gives superior sensitivity and resolution compared to conventional LC.45-49 The coupling of this technique to tandem MS (MS/MS) further enhances specificity and provides an improved signal-to-noise ratio compared with single-stage MS40,41,50,51 and has radically improved the ability to obtain robust and comprehensive BA profiles. Thus, BAs can be measured directly in complex biofluids, and different BA conjugation classes can be distinguished without derivatization.40-44 Although UPLC-MS techniques have been used in a limited number of studies to profile BAs, including the investigation of the modulation of ileal flush BA composition by gut microbiota,10-14 and to explore the toxic effects of carbon tetrachloride and ANIT (36) Marschall, H.-U.; Matern, H.; Wietholtz, H.; Egestad, B.; Matern, S.; Sjo¨vall, J. J. Clin. Invest. 1981, 19928–9. (37) Marschall, H.-U.; Griffiths, W. J.; Zhang, J.; Wietholtz, H.; Matern, H.; Matern, S.; Sjo ¨vall, J. J. Lipid Res. 1994, 35, 1599–1610. (38) Gatti, R.; Roda, A.; Cerre, C.; Bonazzi, D.; Cavrini, V. Biomed. Chromatogr. 1997, 11 (1), 11–15. (39) Nakayama, F.; Nakagaki, M. J. Chromatogr. 1980, 183 (3), 287–293. (40) Burkard, I.; von Eckardstein, A.; Rentsch, K. M. J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 2005, 826 (1-2), 147–59. (41) Ando, M.; Kaneko, T.; Watanabe, R.; Kikuchi, S.; Goto, T.; Iida, T.; Hishinuma, T.; Mano, N.; Goto, J. J. Pharm. Biomed. Anal. 2006, 40 (5), 1179–1186. (42) Mano, N.; Mori, M.; Ando, M.; Goto, T.; Goto, J. J. Pharm. Biomed. Anal. 2006, 40 (5), 1231–1234. (43) Ye, L.; Liu, S.; Wang, M.; Shao, Y.; Ding, M. J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 2007, 860 (1), 10–17. (44) You, J.; Shi, Y.; Zhao, X.; Zhang, H.; Suo, Y.; Yulin, L.; Wang, H.; Sun, J. J. Sep. Sci. 2006, 29 (18), 2837–2846. (45) Plumb, R.; Castro-Perez, J.; Granger, J.; Beattie, I.; Joncour, K.; Wright, A. Rapid Commun. Mass Spectrom. 2004, 18 (19), 2331–2337. (46) Russo, R.; Guillarme, D. T.-T.; Nguyen, D.; Bicchi, C.; Rudaz, S.; Veuthey, J. L. J. Chromatogr. Sci. 2008, 46 (3), 199–208. (47) Wilson, I. D.; Plumb, R.; Granger, J.; Major, H.; Williams, R.; Lenz, E. M. J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 2005, 817 (1), 67–76. (48) Wilson, I. D.; Nicholson, J. K.; Castro-Perez, J.; Granger, J. H.; Johnson, K. A.; Smith, B. W.; Plumb, R. S. J. Proteome Res. 2005, 4, 591–598. (49) Nordstro ¨m, A.; O’Maille, G.; Qin, C.; Siuzdak, G. Anal. Chem. 2006, 78, 3289–3295. (50) Tagliacozzi, D.; Mozzi, A. F.; Casetta, B.; Bertucci, P.; Bernardini, S.; Di Ilio, C.; Urbani, A.; Federici, G. Clin. Chem. Lab. Med. 2003, 41 (12), 1633– 1641. (51) Alnouti, Y.; Csanaky, I. L.; Klaassen, C. D. J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 2008, 873 (2), 209–217.
Analytical Chemistry, Vol. 82, No. 12, June 15, 2010
5283
induced liver failure in rats,52 the number of BAs profiled in these studies did not reflect true BA composition diversity. This may be due to some isobaric BA species not being separated or less common unsaturated BAs not being identified. A multiplatform metabonomic strategy utilizing NMR and UPLC-MS technologies has been employed within the second Consortium on Metabonomic Toxicology (COMET 2) to study galN in more detail. As part of this study, we have developed a series of targeted assays, and here we demonstrate our UPLCMSE 53,54 strategy for the rapid profiling of BAs in biological fluids and illustrate the utility of this approach in the characterization of galN toxicity in a rat model. galN is widely used as a model hepatotoxin, causing reversible liver damage in a dose-dependent manner, with depletion of the hepatic uridine nucleotide pool as the primary toxic effect.55-59 Although many laboratory studies have been conducted, the mechanism of toxicity is not fully understood, particularly with respect to the differential responses elicited in rat models, resulting in “responders” and “nonresponders”.60 Here, the UPLC-MSE 53,54 method is used to characterize serum samples obtained from galN-treated rats and also to investigate the protective effects of glycine against galNmediated liver damage.61-63 EXPERIMENTAL SECTION Reagents and Reference Standards. BA standards were obtained from Steraloids Inc. (London, U.K.). BA stock solutions (1 mg/mL) were prepared in water:methanol (80:20 (v/v)) and stored at -40 °C. Stock solutions were diluted in water to create the required working solutions. Water was Fluka LC-MS CHROMASOLV, and methanol and acetonitrile were Riedel deHahn (all from Sigma, Gillingham, U.K.). Charcoal stripped serum was purchased from Sigma. Calibration Standard Solutions. Five-point calibration standard solutions ranging from 1 ng/mL to 10 µg/mL were prepared by adding appropriate amounts of each BA stock solution into control (charcoal stripped) serum after methanol extraction and stored at -40 °C. (52) Yang, L.; Xiong, A.; He, Y.; Wang, Z.; Wang, C.; Wang, Z.; Li, W.; Yang, L.; Hu, Z. Chem. Res. Toxicol. 2008, 21, 2280–2288. (53) Plumb, R. S.; Johnson, K. A.; Rainville, P.; Smith, B. W.; Wilson, I. D.; CastroPerez, J. M.; Nicholson, J. K. Rapid Commun. Mass Spectrom. 2006, 20 (13), 1989–1994. (54) Bateman, K. P.; Castro-Perez, J.; Wrona, M.; Shockcor, J. P.; Yu, K.; Oballa, R.; Nicoll-Griffith, D. A. Rapid Commun. Mass Spectrom. 2007, 21 (9), 1485– 1496. (55) Decker, K.; Keppler, D. Rev. Physiol. Biochem. Pharmacol. 1974, 71, 77– 106. (56) Keppler, D. O.; Pausch, J.; Decker, K. J. Biol. Chem. 1974, 249, 211–216. (57) Keppler, D.; Frohlich, J.; Reutter, W.; Wieland, O.; Decker, K. FEBS Lett. 1969, 4, 278–80. (58) Keppler, D.; Rudigier, J.; Reutter, W.; Lesch, R.; Decker, K. Hoppe-Seyler Z. Physiol. Chem. 1970, 351, 102–104. (59) Kasravi, F. B.; Wang, L.; Wang, X. D.; Molin, G.; Bengmark, S.; Jeppsson, B. Hepatology 1996, 23, 97–103. (60) Coen, M.; Want, E. J.; Clayton, T. A.; Rohde, C. M.; Hong, Y. S.; Keun, H. C.; Cantor, G. H.; Metz, A. L.; Robertson, D. G.; Reily, M. D.; Holmes, E.; Lindon, J. C.; Nicholson, J. K. J. Proteome Res. 2009, 8, 5175–5187. (61) Wang, B.; Ishihara, M.; Egashira, Y.; Ohta, T.; Sanada, H. Biosci. Biotechnol. Biochem. 1999, 63 (2), 319–322. (62) Coen, M.; Hong, Y. S.; Clayton, T. A.; Rohde, C. M.; Pearce, J. T.; Reily, M. D.; Robertson, D. G.; Holmes, E.; Lindon, J. C.; Nicholson, J. K. J. Proteome Res. 2007, 6, 2711–2719. (63) Stachlewitz, R. F.; Seabra, V.; Bradford, B.; Bradham, C. A.; Rusyn, I.; Germolec, D.; Thurman, R. G. Hepatology 1999, 29, 737–745.
5284
Analytical Chemistry, Vol. 82, No. 12, June 15, 2010
Pooled Serum Samples. A pooled serum sample (representative of the entire sample set) was created by combining 10 µL of each serum sample from the galN treatment study prior to extraction. Serial dilutions (1/2, 1/4, 1/8) were made of this sample and analyzed in triplicate to assess linearity and reproducibility of endogenous BA measurements. The undiluted pooled sample was also injected after every 10 samples to monitor instrument stability. Animals, Treatment, and Sample Collection. Full study details have been reported previously62 and so are given in the Supporting Information (SI). Briefly, 32 male 6 week old Sprague-Dawley rats were housed in a well-ventilated room with water and food supplied ad libitum throughout the 22 day study. On day 21, rats were administered vehicle (0.9% saline, n ) 8, group C), 415 mg/kg galN alone via IP injection (n ) 8, group D1), 5% glycine (via diet, n ) 8, group D2), or a combination of 415 mg/kg galN and 5% glycine (n ) 8, group D3). Groups D2 and D3 received 5% glycine in the diet from day 16. Serum was isolated from blood samples collected at necropsy from the abdominal vena cava and stored at -40 °C pending analysis. A portion of the left lateral liver lobe was also obtained from each animal and stored at -80 °C pending analysis. Clinical Chemistry and Histopathology. (1) Clinical Chemistry Analysis. Serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), and total bilirubin levels were analyzed using a Vitros 950 analyzer (Ortho-Clinical Diagnostics, Rochester, NY). (2) Histological Analysis. Liver samples were fixed in 10% buffered formalin, embedded in paraffin, sectioned, and stained with hematoxylin and eosin. Liver sections were assigned a histopathological necrosis severity score as follows: (0) absence of hepatocellular necrosis, (1) minimal necrosis, (2) mild necrosis, (3) moderate necrosis, and (4) marked necrosis.62 UPLC-MS Analysis of Serum Samples. Serum samples were prepared for UPLC-MS analysis by methanol protein precipitation.64 Cold methanol (150 µL) was added to 50 µL of serum, vortexed for 30 s, incubated at -20 °C for 20 min, centrifuged at 16089g for 10 min, and the supernatant transferred to a clean tube. This supernatant was dried down in a Savant vacuum evaporator (Jencons, West Sussex, U.K.), reconstituted in 100 µL of water, and transferred into 350 µL volume 96-well plates. Metabolite extracts (5 µL) were injected onto a 2.1 × 100 mm (1.7 µm) HSS T3 Acquity column (Waters Corp., Milford, MA) and eluted using a 25 min gradient of 100% A to 100% B (A ) water, 0.1% formic acid; B ) acetonitrile, 0.1% formic acid), with the last 4 min as column re-equilibration. Samples were analyzed using a UPLC system (UPLC Acquity, Waters Ltd., Elstree, U.K.) coupled online to a Q-TOF Premier mass spectrometer (Waters MS Technologies, Ltd., Manchester, U.K.) in negative electrospray mode with a scan range of 50-1000 m/z. BAs ionize strongly in negative mode, producing a prominent [M - H]- ion, and diagnostic fragmentation datasglycine conjugates give rise to a fragment ion at 74 m/z, and taurine conjugates at 79.9, 106, and 124 m/zsallowing conjugates to be differentiated. Capillary voltage was 2.4 kV, sample cone was 35 V, desolvation temperature 350 °C, source temperature 120 °C, and desolvation gas flow 900 L/h. Mass (64) Want, E. J.; O’Maille, G.; Smith, C. A.; Brandon, T. R.; Uritboonthai, W.; Qin, C.; Trauger, S. A.; Siuzdak, G. Anal. Chem. 2006, 78, 743–752.
spectrometric conditions were optimized through direct infusion of available BA standards. The Q-Tof Premier was operated in V optics mode, with a data acquisition rate of 0.1 s and a 0.01 s interscan delay. Leucine enkephalin (m/z 556.2771) was used as the lockmass; a solution of 200 pg/µL (50:50 ACN: H2O) was infused into the instrument at 3 µL/min via an auxiliary sprayer. Data were collected in centroid mode with a scan range of 50-1000 m/z, with lockmass scans collected every 15 s and averaged over 3 scans to perform mass correction. MSE was performed on all samples, where data were collected at both low (5 V) and high (50 V) collision energies, in order to obtain fragmentation data simultaneously.53,54 Method Validation. After assay optimization, the following method validation procedures were employed on the basis of the FDA guidelines for Bioanalytical Method Validation.65 Method Linearity. The BA calibration range was established using the following calibration standard concentrations in control serum: 1 ng/mL, 10 ng/mL, 100 ng/mL, 1 µg/mL, and 10 µg/ mL. Calibration curve linearity was evaluated using three consecutively prepared and analyzed batches of standards, with each sample being analyzed six times in the separate runs. Standard curves were constructed by least-squares linear regression analysis using the BA peak area as calculated using the ApexTrack2 function in Masslynx software (Waters Corp.) versus the concentration of standard. The control (charcoal stripped) serum into which the BA standards were spiked was evaluated using the BA assay and found to have no detectable levels of the BAs being profiled. Precision and Accuracy. Six replicates of each calibration standard in three different serum preparations (0.1 µg/mL, 1 µg/ mL, and 10 µg/mL BA standard concentrations) were analyzed on three separate days to determine the intra- and inter-day accuracy and precision. Coefficients of variation (% CV) of peak area intensities were calculated for each conjugation group of BAs (unconjugated, glycine-conjugated, and taurine-conjugated). Intra- and Inter-assay Precision of Endogenous BAs. The undiluted pooled sample and the three dilutions (1/2, 1/4, 1/8) were run in triplicate in a random order throughout the sample set on 2 separate days to assess intra- and inter-assay precision. Linearity of endogenous BAs was also assessed, including BAs for which no authentic standard was available. Recovery. Extraction procedure efficiency was determined through the analysis of bile acid standards spiked into charcoal stripped serum pre- and post-methanol extraction. (1) Pre-extraction. BAs were added to 50 µL of serum to give concentrations of 1 and 5 µg/mL and extracted six times as described above. (2) Post-extraction. Serum was extracted six times as described above, and BAs were added during sample resuspension to give a final concentration of 1 or 5 µg/mL. Average BA peak areas were compared between pre- and post-extraction methods to measure extraction efficiency. Data Analysis. UPLC-MS data sets were preprocessed via peak peaking and alignment before multivariate statistical analysis, using a combination of software to maximize the metabolic (65) U.S. Department of Health and Human Services Food and Drug Administration, Center for Drug Evaluation and Research (CDER), Center for Veterinary Medicine (CVM), May 2001; http://www.fda.gov/CDER/GUIDANCE/4252fnl. htm.
information obtained. MarkerLynx application manager in Masslynx software (Waters Corp.) and the freeware package XCMS66 were used to generate “marker tables” comprising m/z, RT, and intensity (peak area) values for each variable in every sample, which were exported into SIMCA-P (Umetrics) for further multivariate analysis. Marker tables contained isotopes and adducts, e.g., Na+ as well as in-source created fragments. A student’s t test was used to investigate differences between the classes in clinical chemistry and BA measurements. The following analytical approach was taken to maximize data interpretation. (1) Comparison of Relative BA Levels. From comparison with standards where possible, the retention time and m/z [M - H](m/z_RT pair) of each serum BA were recorded. The identity of several unsaturated BAs could not be confirmed using standards and so was postulated on the basis of retention time, m/z, and fragmentation information. BA intensity (peak area) values were used to calculate relative BA proportions in each sample group. (2) Principal Components Analysis (PCA) of BA Profile Data. Principal components analysis (PCA) was used to elucidate which BAs were collectively most important in separating the treatment groups. Each m/z_RT pair intensity value was used as a descriptor for PCA. Two-dimensional PC scores and loading plots of PC1 versus PC2 were constructed and conclusions drawn from these with respect to the identity and relative proportions of key BAs. (3) Correlation with Histopathology. (i) Box and Whisker Plots. Box and whisker plots were generated in MATLAB (The Mathworks, Natick, MA) to show the relationship of clinical chemistry measurements (AST, ALT, and bilirubin) and BA measurements (taurine, glycine, and unconjugated BAs) with the scored histopathology data. (ii) Correlation Analysis. A Pearson’s rank correlation coefficient (r) was calculated between the BA classes and (i) clinical chemistry and (ii) histopathology measurements. RESULTS AND DISCUSSION Method Validation. (1) Linearity. A linear calibration curve was constructed for each BA using the regression of the peak area versus the concentration of the BA standard. Correlation was linear over a 103 dynamic range covering 10 ng/mL to 10 µg/ mL. All BAs gave excellent linear responses, with coefficients of determination (R2) in the range of 0.9935-0.9998 (Table 1). (2) Precision and Accuracy. Mean intra-day accuracy was 4-14% over all BA conjugation classes, with an overall mean CV of 9%, and mean inter-day accuracy was 3-16%, with an overall mean CV of 9% (Table 2A). Triplicate analyses of endogenous BAs in the pooled serum sample were precise and accurate with CVs from 4 to 16% and a good linear response observed over an 8-fold dilution range (Table 2B). (3) Retention Time Reproducibility. A minimal shift in retention time both within and between analyses is important for confident metabolite identification. An important attribute of UPLC is its high reproducibility compared to conventional HPLC, crucial in metabolite profiling studies. The mean retention time variation of the BAs was
