- Email: [email protected]
MS method for the selective determination of paracetamol metabolites in mouse urine
Available online at www.sciencedirect.com
ANALYTICAL BIOCHEMISTRY Analytical Biochemistry 374 (2008) 106–111 www.elsevier.com/locate/yabio
Development of an HPLC–MS/MS method for the selective determination of paracetamol metabolites in mouse urine A.K. Hewavitharana b
a,*
, S. Lee b, P.A. Dawson b, D. Markovich b, P.N. Shaw
a
a School of Pharmacy, University of Queensland, Brisbane 4072, Australia School of Biomedical Sciences, University of Queensland, Brisbane 4072, Australia
Received 24 August 2007 Available online 17 November 2007
Abstract An HPLC–MS/MS method has been developed for the selective quantitative analysis of paracetamol and its two major metabolites. The use of tandem MS enabled the detection and quantitation of metabolites in small sample sizes with high sensitivity and selectivity. Isocratic elution using acetonitrile and water containing formic acid combined with electrospray–tandem MS enabled the separation and accurate quantitation of each analyte and the internal standard 3-acetamidophenol. The on-column limits of detection for paracetamol, paracetamol sulfate, and paracetamol glucuronide were 2.4, 1.2, and 1.2 pmol, respectively. The method was applied to quantitate paracetamol and its metabolites in mouse urine. It is highly specific, sensitive, and easily adaptable to measure these analytes in biological fluids of other animals. 2007 Elsevier Inc. All rights reserved. Keywords: HPLC; HPLC–MS; APAP; Acetaminophen; Metabolites; Tandem MS
Paracetamol or acetaminophen (N-acetyl-p-aminophenol [APAP])1 is a widely used analgesic. Study of paracetamol metabolism is important in toxicological and pharmacokinetic studies of the drug. Paracetamol glucuronide (PG) and paracetamol sulfate (PS) are major metabolites of paracetamol that are detected in urine (Fig. 1). It has been well documented that glucuronidation and sulfation of the phenolic group of APAP are the major detoxification pathways in most mammalian species [1]. Commonly available HPLC methods for the quantitation of paracetamol and its metabolites employ nonselective UV spectrophotometric detection [2–4]. Unlike UV, mass spectrometric detection is specific for each analyte; there-
*
Corresponding author. Fax: +61 7 3365 1688. E-mail address: [email protected] (A.K. Hewavitharana). 1 Abbreviations used: APAP, N-acetyl-p-aminophenol; PG, paracetamol glucuronide; PS, paracetamol sulfate; MRM, multiple reaction monitoring; IS, internal standard; PTFE, polytetrafluoroethene; PC, paracetamol cysteinate; PM, paracetamol mercapturate; TIC, total ion chromatogram. 0003-2697/$ - see front matter 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2007.11.011
fore, the risk of overestimation of the analytes due to coelution with other compounds is eliminated. Tandem MS (LC–MS/MS) mode with multiple reaction monitoring (MRM) further enhances the accuracy of the method because the MRM response is due to the presence of both the analyte ion and its specific fragment. Furthermore, the LC–MS/MS mode is far superior to the LC–MS mode in terms of sensitivity and, thus, is well suited for quantitative analysis. In this study, the superior sensitivity of the technique (vs. the commonly used LC–UV) was required to measure the low levels of metabolites found in small sample sizes of mouse urine. There have been several studies where LC–MS (single quadrupole mode) was used for the quantitation of paracetamol [5–7]. However, most reported studies of liquid chromatographic determination of paracetamol metabolites in biological fluids have been only qualitative using the single quadrupole mode (LC–MS) [8–11]. The only quantitative analysis of paracetamol and metabolites using mass spectrometric detection also employed only LC–MS after derivatizing the metabolites to increase the sensitivity
Selective determination of paracetamol metabolites / A.K. Hewavitharana et al. / Anal. Biochem. 374 (2008) 106–111
107
Fig. 1. Structures of paracetamol and its metabolites.
of the single quadrupole detection [12]. Although LC–MS/ MS has been used for the determination of paracetamol [13–15], to date it has not been used for the determination of paracetamol metabolites. We present here an LC–MS/ MS method for the quantitation of paracetamol and its metabolites using an isomeric form of paracetamol, 3-acetamidophenol, as the internal standard (IS).
of a similar size also gave adequate separation of all analytes (data not shown). An API 3000 tandem mass spectrometer equipped with a turbo ion spray interface and supported by Analyst 1.4 software (Applied Biosystems, Foster City, CA, USA) was used to detect the separated compounds and process all data. Materials
Materials and methods Instrumentation The separation of compounds was carried out using an Agilent binary HPLC system consisting of an Agilent 1100 LC pump and an Agilent 1100 well plate autosampler. ˚ (2.1 · 50 mm, 5 lM) Although a Cogent phenyl 100-A HPLC column (Microsolv Technology, Long Branch, NJ, USA) was used for this study (because the analytes retained better in the phenyl column), we found that a C18 column
Paracetamol, PG, and 3-acetamidophenol were purchased from Sigma (St. Louis, MO, USA). PS was obtained from Toronto Research Chemicals (North York, ON, Canada). All solvents used were of HPLC grade. Methods Mobile phase Eluent A was prepared by adding 0.1% (v/v) formic acid to water, followed by filtering through a polytetrafluoro-
108
Selective determination of paracetamol metabolites / A.K. Hewavitharana et al. / Anal. Biochem. 374 (2008) 106–111
ethene (PTFE) 0.45-lm filter (Millipore, Bedford, MA, USA). Eluent B was prepared in the same manner using acetonitrile instead of water. Standard solutions Stock standard solutions of all compounds were prepared in water and stored at 20 C. A combination standard containing paracetamol (0.4 mmol/L), PG (0.2 mmol/ L), PS (0.02 mmol/L), and the IS (0.1 mmol/L) was prepared by adding appropriate amounts of standard compounds to blank urine (pooled urine from untreated control mice) that was diluted to the same extent as were the samples. Sample preparation Studies were performed on male mice with a mixed genetic background (129Sv and C57BL/6 J). Mice were weaned at 3 weeks of age and then were fed a standard rodent chow (cat. no. AIN93G, Glen Forrest Stockfeeders, Glen Forrest, WA, Australia) and water ad libitum. Paracetamol was dissolved in saline and filter sterilized. At approximately 3 months of age, mice received an intraperitoneal injection of saline (control) or a 250-mg/kg dose of paracetamol [16]. Urine was collected 2 h after saline or paracetamol administration and then stored at 80 C until needed. All experiments conformed to the guidelines of the University of Queensland Animal Ethics Committee. Immediately prior to analysis, urine samples were thawed and 10 ll of 1 mmol/L IS was added to 10 ll of urine and then diluted 10-fold by adding 80 ll of eluent A. Each sample was then mixed by vortexing and centrifuged at 10,000 g for 5 min before injection. Samples were prepared in duplicate. Chromatographic conditions Chromatography was performed at ambient temperature. The mobile phase flow rate through the column was 200 ll/min with 15 ll/min introduced into the ion spray mass spectrometer. A splitter and suitable lengths of tubing were used to split the mobile phase flow into the appropriate ratio. Following injection of each sample or standard, analytes were separated with an isocratic elution using 3% eluent A for 10 min. Column was then flushed with 100% eluent B by raising the eluent strength to 100% eluent B over 3 min and then holding at that composition for 2 min. The composition was then changed to the original composition of 3% eluent B over 5 min, and the column was reequilibrated for 10 min with 3% eluent A prior to injection of each sample or standard. MRM positive ion mode MS The detector response in MRM is due to a specific transition of molecular ion fi fragment. The fragmentations of 152 fi 110 m/z ion (for paracetamol, IS), 232 fi 152 m/z ion (for PS), and 328 fi 152 m/z ion (for PG) were monitored for each chromatographic run. The following are
the parameters optimized for each compound to obtain the highest possible sensitivity for each. An ion spray voltage of 5000 V and an entrance potential of 10 V were used in all analyses. Orifice/Declustering potentials were 36 V (paracetamol, IS), 46 V (PS), and 31 V (PG). Ring/ Focusing potentials were 190 V (paracetamol, IS), 250 V (PS), and 160 V (PG). Collision energies were 23 V (paracetamol, IS), 21 V (PS), and 15 V (PG). Collision exit potentials were 6 V (paracetamol, IS) and 10 V (PS and PG). Curtain gas, nebulizer gas, and the collision gas flows were maintained at 8 (in arbitrary units used in the instrument). The temperature of the ion spray was kept at ambient. A dwell time of 1 s was used for all transitions. The resolution of both Q1 and Q3 was 1 amu. Quantitative analysis Quantitative analysis was performed by five-point calibration, covering the expected concentrations of analytes in the extracts, using the combination standard. An injection volume of 30 ll was used for all samples. Quantitation was performed using the IS method to compensate for the sample losses during extraction and analysis. Quantitation of each analyte was performed at its own specific MRM response on duplicates of each sample. Results and discussion Standard solutions used for the construction of calibration curves were prepared by adding various known amounts of analytes to the urine of untreated (control) animals that was diluted to the same extent as the samples. Therefore, any influence of matrix, including ion suppression effects in the electrospray process, are compensated because samples and standards contained the same matrix. To match the pH of the eluent to that of the standards and samples, all dilutions were made with the mobile phase. Attempts to perform negative ion MS, using ammonium acetate in place of formic acid in solution, were unsuccessful because there was no response for the paracetamol. In positive ion MS, the protonated molecular ions of paracetamol, PG, PS, and the IS were produced by the protonation of the N-acetyl group at the N position producing secondary ammonium groups. The fragmentation of PG and PS produced 152 m/z ion, which is the same as the protonated molecular ion of paracetamol. The fragmentation of paracetamol was due to the elimination of acetyl fragment of the N-acetyl group to produce a primary 110 m/z ammonium ion. The specificity of the method to each analyte was achieved by matching the retention time of each analyte in the sample to that in the standard and by detecting the protonated molecular ion plus a specific fragment (152 or 110 m/z) for each analyte. Further confirmation was achieved by spiking the samples with the standard compounds, followed by monitoring the growth of each peak. The urine of saline-treated mice (blank) yielded no detectable peaks on the chromatograph, whereas urine
Selective determination of paracetamol metabolites / A.K. Hewavitharana et al. / Anal. Biochem. 374 (2008) 106–111
samples from paracetamol-treated mice showed peaks corresponding to paracetamol and its metabolites (Fig. 2). As seen in Fig. 2, both PG and PS produced two MRM peaks at the same retention time as the main peak: one due to the monitored MRM transition (328 fi 152 m/z ion for PG and 232 fi 152 m/z ion for PS) and another due to an additional secondary transition of 152 fi 110 m/z ion (for both PG and PS because both produce 152 m/z ion from the primary transition) . Because the method is set up to detect 152 fi 110 m/z ion (as the primary transition for paracetamol), the secondary transition of PG and PS (which is the same as the primary transition of paracetamol) is also detected at their respective retention times (second trace of PG and PS peaks in Fig. 2). This gave an additional degree of confirmation for PG and PS (in addition to retention time and the primary MRM transition), increasing the specificity of the method for these two analytes. However, for PS two separate urinary PS peaks were observed at different retention times (PS1 and PS2). Because the standard chromatogram contained only one peak for PS and the chromatogram of the blank urine showed no peaks, the additional peak most likely is the result of paracetamol metabolism in mice. The appearance of two peaks was not due to a chromatographic problem such as peak splitting (due to column overloading) because the PS in the standard solution did not appear as a split peak even at very high concentrations. Spiking confirmed that PS1 was the peak due to the standard compound
Fig. 2. Representative chromatogram of a mouse urine sample at 2 h after paracetamol administration (250 mg/kg).
109
PS, in which the sulfate group is in the para position to the N-acetyl group (shown as PS in Fig. 1). One way to explain the appearance of PS2 is that it may have been due to a secondary transition of a metabolite that was not detected by the method described here. The metabolites suspected were paracetamol cysteinate (PC, MW = 270) and paracetamol mercapturate (PM, MW = 312) (Fig. 1), with the suspected primary transitions 271 fi 232 m/z ion and 313 fi 232 m/z ion, respectively. If this is the case, then the secondary transition of 232 fi 152 m/z ion could have caused the PS2 peak. When the primary transitions 271 fi 232 m/z ion and 313 fi 232 m/z ion were monitored for urine samples, no peaks were observed, implying that the PS2 peak was not due to the secondary transition of either PC or PM. For a thorough investigation, a total ion chromatogram (TIC) was run for a sample within the range of 200 to 350 m/z. An attempt was made to extract a variety of suspected metabolite ions from TIC: 313 (PM), 271 (PC), and 232 (paracetamol-4-sulfate [N-acetyl4-aminophenol sulfate] or paracetamol-3-sulfate [N-acetyl3-aminophenol sulfate]). In addition to these, we attempted to extract three further putative metabolites from the TIC: paracetamol N-sulfate (N-acetyl-N-sulfonyl-4-aminophenol), 248 (N-acetyl-3-sulfonyl-4-aminophenol) (II), and 263 (N-acetyl-3-sulfonyl-4-aminomethoxyphenol) (III) (Fig. 1). The extracted ion chromatogram from TIC, however, showed only one peak that was PC. The other metabolites were not in sufficiently high concentrations to be detected in TIC mode. To further confirm that the PC is not responsible for producing the PS2 peak, a fragment ion (MS2) chromatogram was run for 271 (PC) ion in the sample. This produced its major fragment at 152 (same as PG and PS) rather than at 232. Furthermore, the MS2 chromatogram of the sample for 232 ion produced two peaks, at the retention times of PS1 and PS2, with identical fragment ion spectra (with 152 m/z as the major fragment). This chromatogram and spectra are shown in Fig. 3. In conclusion, the PS2 peak is produced by an isomer of PS that has the same molecular weight. MS is limited to provide information about the position of the sulfate group. Possible compounds are paracetamol-3-sulfate (i.e., sulfate group is in the meta position to the N-acetyl group) and paracetamol N-sulfate (I). Standard compounds are not commercially available for either of these compounds; therefore, we were unable to confirm the identity of PS2. In the literature, there was no mention of the formation of an additional isomer of PS during paracetamol metabolism, and we suspect this is due to the nonspecific nature of the detection (by UV) commonly used to monitor the metabolic products of paracetamol. Because there were many additional peaks (other than the peaks eluting at the retention times of the metabolite standards) with UV detection, an additional peak such as PS2 would have been ignored as a matrix peak. In this study, the use of tandem MS for the detection of paracetamol metabolites provides three degrees of specificity (retention time, MS1, and MS2) for each metabolite, whereas the UV
110
Selective determination of paracetamol metabolites / A.K. Hewavitharana et al. / Anal. Biochem. 374 (2008) 106–111
Fig. 3. Chromatogram of a mouse urine sample after paracetamol administration showing the fragmentation of 232 ion detected in MS2 mode. The fragmentation patterns of the PS1 and PS2 peaks are shown in the middle and lower panels, respectively.
method relies on only one degree of specificity (retention time). Notarianni and coworkers [17] detected a second peak for PS, and it was identified as one of the minor metabolites (paracetamol-3-sulfate). The order of elution is also consistent with our results where the para-substituted sulfate elutes before the meta substitute. This trend is also evident with the order of elution of paracetamol, which is the para substitute, and the IS, which is the meta substitute. We also found that the IS response in tandem MS is more sensitive than the paracetamol response. If the sulfates also follow the same trend, then it is likely that PS2, although detected as a significant peak, is likely to occur at very small concentrations compared with PS1. Taken together, the current and previous [17] studies suggest that the PS2 peak is most likely to be paracetamol-3sulfate. Quantitation of the metabolites in urine samples was carried out using paracetamol, PG, and PS standards and the structurally similar IS 3-acetamidophenol. Calibration curves consisted of five points for each analyte and were linear from 0 to 1.2 mmol/L for paracetamol (R2 = 0.998), from 0 to 0.6 mmol/L for PG (R2 = 0.996), and
from 0 to 0.06 mmol/L for PS (R2 = 0.969) with 30-ll injection volumes. The ranges of concentrations observed in mouse urine 2 h after intraperitoneal injection of 250 mg/ kg paracetamol (after correction for dilution) were 1.05 to 2.68 mmol/L for paracetamol, 1.74 to 2.48 mmol/L for PG, and 0.05 to 0.13 mmol/L for PS. The on-column limit of detection (LOD) determined, using standards prepared by spiking urine with each analyte and estimated as three times noise, were 2.4 pmol for paracetamol and 1.2 pmol for PG and PS. Considering the sample dilution step and the volume injected (30 ll), these equate to 0.66 lmol/L of paracetamol and 0.33 lmol/L of PG and PS in urine. These limits are lower than those reported previously [2] for UV detection: 2.2 lmol/L (for paracetamol), 0.52 lmol/L (for PG), and 1.4 lmol/L (for PS). The repeatability values (n = 8), determined by analyzing multiple samples of the same batch on the same day and expressed as percentage standard deviations, were 2.08% for paracetamol, 1.62% for PG, and 7.40% for PS. The method has high specificity due to tandem MS detection, and it is consistent and fast with a total run time of 30 min, including a column wash and a 10-min column reequilibration.
Selective determination of paracetamol metabolites / A.K. Hewavitharana et al. / Anal. Biochem. 374 (2008) 106–111
Conclusion The LC–MS/MS technique was applied to detect and quantify paracetamol metabolites in biological fluids. The unique attributes of this technique enabled the detection and quantitation of metabolites with high specificity and sensitivity. The method may be adapted for the determination of paracetamol metabolites in other biological fluids of human and other animals.
[7]
[8]
[9]
Acknowledgments This study was funded in part by NHMRC and ARC grants to D.M. S.L. is a recipient of the University of Queensland Confirmation Scholarship.
[10]
[11]
References [1] J.R. Mitchell, D.J. Jollow, S.W. Potter, D.C. Davis, J.R. Gillette, B.B. Brodie, Acetaminophen-induced hepatic necrosis: I. Role of drug metabolism, J. Pharmacol. Exp. Ther. 187 (1973) 185–194. [2] M.V. Vertzoni, H.A. Archontaki, P. Galanopoulou, Development and optimization of a reversed-phase high-performance liquid chromatographic method for the determination of acetaminophen and its major metabolites in rabbit plasma and urine after a toxic dose, J. Pharm. Biomed. Anal. 32 (2003) 487–493. [3] A. Di Girolamo, W.M. O’Neill, I.W. Wainer, A validated method for the determination of paracetamol and its glucuronide and sulphate metabolites in the urine of HIV+ /AIDS patients using wavelengthswitching UV detection, J. Pharm. Biomed. Anal. 17 (1998) 1191– 1197. [4] A.W. Abu-Qare, M.B. Abou-Donia, A validated HPLC method for the determination of pyridostigmine bromide, acetaminophen, acetylsalicylic acid, and caffeine in rat plasma and urine, J. Pharm. Biomed. Anal. 26 (2001) 939–947. [5] T. Matsumoto, T. Sano, T. Matsuoka, M. Aoki, Y. Maeno, M. Nagao, Simultaneous determination of carisoprodol and acetaminophen in an attempted suicide by liquid chromatography–mass spectrometry with positive electrospray ionization, J. Anal. Toxicol. 27 (2003) 118–122. [6] A. Panusa, G. Multari, G. Incarmato, L. Gagliardi, High-performance liquid chromatography analysis of anti-inflammatory pharmaceuticals with ultraviolet and electrospray–mass spectrometry
[12]
[13]
[14]
[15]
[16]
[17]
111
detection in suspected counterfeit homeopathic medicinal products, J. Pharm. Biomed. 43 (2007) 1221–1227. I.R. Miksa, M.R. Cummings, R.H. Poppenga, Multi-residue determination of anti-inflammatory analgesics in sera by liquid chromatography–mass spectrometry, J. Anal. Toxicol. 29 (2005) 95–104. S. La, H.H. Yoo, D.H. Kim, Liquid chromatography–mass spectrometric analysis of urinary metabolites and their pattern recognition for the prediction of drug-induced hepatotoxicity, Chem. Res. Toxicol. 18 (2005) 1887–1896. L.D. Betowski, W.A. Korfmacher, J.O. Lay Jr., D.W. Potter, J.A. Hinson, Direct analysis of rat bile for acetaminophen and two of its conjugated metabolites via thermospray liquid chromatography mass spectrometry, Biomed. Environ. Mass Spectrom. 14 (1987) 705–709. B.L. Ackermann, J.T. Watson, J.F. Newton Jr., J.B. Hook, W.E. Braselton Jr., Application of fast atom bombardment mass spectrometry to biological samples: Analysis of urinary metabolites of acetaminophen, Biomed. Mass Spectrom. 11 (1984) 502–511. T.A. Getek, W.A. Korfmacher, T.A. McRae, J.A. Hinson, Utility of solution electrochemistry mass spectrometry for investigating the formation and detection of biologically important conjugates of acetaminophen, J. Chromatogr. 474 (1989) 245–256. M. Ohta, N. Kawakami, S. Yamato, K. Shimada, Analysis of acetaminophen glucuronide conjugate accompanied by adduct ion production by liquid chromatography–atmospheric pressure chemical ionization–mass spectrometry, J. Pharm. Biomed. 30 (2003) 1759– 1764. O.Q.P. Yin, S.S.L. Lam, M.S.S. Chow, Simultaneous determination of paracetamol and dextropropoxyphene in human plasma by liquid chromatography/tandem mass spectrometry: Application to clinical bioequivalence studies, Rapid Commun. Mass Spectrom. 19 (2005) 767–774. X. Chen, J. Huang, Z. Kong, D. Zhong, Sensitive liquid chromatography–tandem mass spectrometry method for the simultaneous determination of paracetamol and guaifenesin in human plasma, J. Chromatogr. B 817 (2005) 263–269. C. Celma, J.A. Allue, J. Prunonosa, C. Peraire, R. Obach, Simultaneous determination of paracetamol and chlorpheniramine in human plasma by liquid chromatography–tandem mass spectrometry, J. Chromatogr. A 870 (2000) 77–86. S. Lee, P.A. Dawson, A.K. Hewavitharana, P.N. Shaw, D. Markovich, Disruption of NaS1 sulfate transport function in mice leads to enhanced acetaminophen-induced hepatotoxicity, Hepatology 43 (2006) 1241–1247. L.J. Notarianni, H.G. Oldham, P.N. Bennett, C.C.B. Southgate, R.T. Parfitt, Epoxides from paracetamol: A possible explanation for paracetamol toxicity, Adv. Exp. Med. Biol. 136 (1982) 1077–1083.
ANALYTICAL BIOCHEMISTRY Analytical Biochemistry 374 (2008) 106–111 www.elsevier.com/locate/yabio
Development of an HPLC–MS/MS method for the selective determination of paracetamol metabolites in mouse urine A.K. Hewavitharana b
a,*
, S. Lee b, P.A. Dawson b, D. Markovich b, P.N. Shaw
a
a School of Pharmacy, University of Queensland, Brisbane 4072, Australia School of Biomedical Sciences, University of Queensland, Brisbane 4072, Australia
Received 24 August 2007 Available online 17 November 2007
Abstract An HPLC–MS/MS method has been developed for the selective quantitative analysis of paracetamol and its two major metabolites. The use of tandem MS enabled the detection and quantitation of metabolites in small sample sizes with high sensitivity and selectivity. Isocratic elution using acetonitrile and water containing formic acid combined with electrospray–tandem MS enabled the separation and accurate quantitation of each analyte and the internal standard 3-acetamidophenol. The on-column limits of detection for paracetamol, paracetamol sulfate, and paracetamol glucuronide were 2.4, 1.2, and 1.2 pmol, respectively. The method was applied to quantitate paracetamol and its metabolites in mouse urine. It is highly specific, sensitive, and easily adaptable to measure these analytes in biological fluids of other animals. 2007 Elsevier Inc. All rights reserved. Keywords: HPLC; HPLC–MS; APAP; Acetaminophen; Metabolites; Tandem MS
Paracetamol or acetaminophen (N-acetyl-p-aminophenol [APAP])1 is a widely used analgesic. Study of paracetamol metabolism is important in toxicological and pharmacokinetic studies of the drug. Paracetamol glucuronide (PG) and paracetamol sulfate (PS) are major metabolites of paracetamol that are detected in urine (Fig. 1). It has been well documented that glucuronidation and sulfation of the phenolic group of APAP are the major detoxification pathways in most mammalian species [1]. Commonly available HPLC methods for the quantitation of paracetamol and its metabolites employ nonselective UV spectrophotometric detection [2–4]. Unlike UV, mass spectrometric detection is specific for each analyte; there-
*
Corresponding author. Fax: +61 7 3365 1688. E-mail address: [email protected] (A.K. Hewavitharana). 1 Abbreviations used: APAP, N-acetyl-p-aminophenol; PG, paracetamol glucuronide; PS, paracetamol sulfate; MRM, multiple reaction monitoring; IS, internal standard; PTFE, polytetrafluoroethene; PC, paracetamol cysteinate; PM, paracetamol mercapturate; TIC, total ion chromatogram. 0003-2697/$ - see front matter 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2007.11.011
fore, the risk of overestimation of the analytes due to coelution with other compounds is eliminated. Tandem MS (LC–MS/MS) mode with multiple reaction monitoring (MRM) further enhances the accuracy of the method because the MRM response is due to the presence of both the analyte ion and its specific fragment. Furthermore, the LC–MS/MS mode is far superior to the LC–MS mode in terms of sensitivity and, thus, is well suited for quantitative analysis. In this study, the superior sensitivity of the technique (vs. the commonly used LC–UV) was required to measure the low levels of metabolites found in small sample sizes of mouse urine. There have been several studies where LC–MS (single quadrupole mode) was used for the quantitation of paracetamol [5–7]. However, most reported studies of liquid chromatographic determination of paracetamol metabolites in biological fluids have been only qualitative using the single quadrupole mode (LC–MS) [8–11]. The only quantitative analysis of paracetamol and metabolites using mass spectrometric detection also employed only LC–MS after derivatizing the metabolites to increase the sensitivity
Selective determination of paracetamol metabolites / A.K. Hewavitharana et al. / Anal. Biochem. 374 (2008) 106–111
107
Fig. 1. Structures of paracetamol and its metabolites.
of the single quadrupole detection [12]. Although LC–MS/ MS has been used for the determination of paracetamol [13–15], to date it has not been used for the determination of paracetamol metabolites. We present here an LC–MS/ MS method for the quantitation of paracetamol and its metabolites using an isomeric form of paracetamol, 3-acetamidophenol, as the internal standard (IS).
of a similar size also gave adequate separation of all analytes (data not shown). An API 3000 tandem mass spectrometer equipped with a turbo ion spray interface and supported by Analyst 1.4 software (Applied Biosystems, Foster City, CA, USA) was used to detect the separated compounds and process all data. Materials
Materials and methods Instrumentation The separation of compounds was carried out using an Agilent binary HPLC system consisting of an Agilent 1100 LC pump and an Agilent 1100 well plate autosampler. ˚ (2.1 · 50 mm, 5 lM) Although a Cogent phenyl 100-A HPLC column (Microsolv Technology, Long Branch, NJ, USA) was used for this study (because the analytes retained better in the phenyl column), we found that a C18 column
Paracetamol, PG, and 3-acetamidophenol were purchased from Sigma (St. Louis, MO, USA). PS was obtained from Toronto Research Chemicals (North York, ON, Canada). All solvents used were of HPLC grade. Methods Mobile phase Eluent A was prepared by adding 0.1% (v/v) formic acid to water, followed by filtering through a polytetrafluoro-
108
Selective determination of paracetamol metabolites / A.K. Hewavitharana et al. / Anal. Biochem. 374 (2008) 106–111
ethene (PTFE) 0.45-lm filter (Millipore, Bedford, MA, USA). Eluent B was prepared in the same manner using acetonitrile instead of water. Standard solutions Stock standard solutions of all compounds were prepared in water and stored at 20 C. A combination standard containing paracetamol (0.4 mmol/L), PG (0.2 mmol/ L), PS (0.02 mmol/L), and the IS (0.1 mmol/L) was prepared by adding appropriate amounts of standard compounds to blank urine (pooled urine from untreated control mice) that was diluted to the same extent as were the samples. Sample preparation Studies were performed on male mice with a mixed genetic background (129Sv and C57BL/6 J). Mice were weaned at 3 weeks of age and then were fed a standard rodent chow (cat. no. AIN93G, Glen Forrest Stockfeeders, Glen Forrest, WA, Australia) and water ad libitum. Paracetamol was dissolved in saline and filter sterilized. At approximately 3 months of age, mice received an intraperitoneal injection of saline (control) or a 250-mg/kg dose of paracetamol [16]. Urine was collected 2 h after saline or paracetamol administration and then stored at 80 C until needed. All experiments conformed to the guidelines of the University of Queensland Animal Ethics Committee. Immediately prior to analysis, urine samples were thawed and 10 ll of 1 mmol/L IS was added to 10 ll of urine and then diluted 10-fold by adding 80 ll of eluent A. Each sample was then mixed by vortexing and centrifuged at 10,000 g for 5 min before injection. Samples were prepared in duplicate. Chromatographic conditions Chromatography was performed at ambient temperature. The mobile phase flow rate through the column was 200 ll/min with 15 ll/min introduced into the ion spray mass spectrometer. A splitter and suitable lengths of tubing were used to split the mobile phase flow into the appropriate ratio. Following injection of each sample or standard, analytes were separated with an isocratic elution using 3% eluent A for 10 min. Column was then flushed with 100% eluent B by raising the eluent strength to 100% eluent B over 3 min and then holding at that composition for 2 min. The composition was then changed to the original composition of 3% eluent B over 5 min, and the column was reequilibrated for 10 min with 3% eluent A prior to injection of each sample or standard. MRM positive ion mode MS The detector response in MRM is due to a specific transition of molecular ion fi fragment. The fragmentations of 152 fi 110 m/z ion (for paracetamol, IS), 232 fi 152 m/z ion (for PS), and 328 fi 152 m/z ion (for PG) were monitored for each chromatographic run. The following are
the parameters optimized for each compound to obtain the highest possible sensitivity for each. An ion spray voltage of 5000 V and an entrance potential of 10 V were used in all analyses. Orifice/Declustering potentials were 36 V (paracetamol, IS), 46 V (PS), and 31 V (PG). Ring/ Focusing potentials were 190 V (paracetamol, IS), 250 V (PS), and 160 V (PG). Collision energies were 23 V (paracetamol, IS), 21 V (PS), and 15 V (PG). Collision exit potentials were 6 V (paracetamol, IS) and 10 V (PS and PG). Curtain gas, nebulizer gas, and the collision gas flows were maintained at 8 (in arbitrary units used in the instrument). The temperature of the ion spray was kept at ambient. A dwell time of 1 s was used for all transitions. The resolution of both Q1 and Q3 was 1 amu. Quantitative analysis Quantitative analysis was performed by five-point calibration, covering the expected concentrations of analytes in the extracts, using the combination standard. An injection volume of 30 ll was used for all samples. Quantitation was performed using the IS method to compensate for the sample losses during extraction and analysis. Quantitation of each analyte was performed at its own specific MRM response on duplicates of each sample. Results and discussion Standard solutions used for the construction of calibration curves were prepared by adding various known amounts of analytes to the urine of untreated (control) animals that was diluted to the same extent as the samples. Therefore, any influence of matrix, including ion suppression effects in the electrospray process, are compensated because samples and standards contained the same matrix. To match the pH of the eluent to that of the standards and samples, all dilutions were made with the mobile phase. Attempts to perform negative ion MS, using ammonium acetate in place of formic acid in solution, were unsuccessful because there was no response for the paracetamol. In positive ion MS, the protonated molecular ions of paracetamol, PG, PS, and the IS were produced by the protonation of the N-acetyl group at the N position producing secondary ammonium groups. The fragmentation of PG and PS produced 152 m/z ion, which is the same as the protonated molecular ion of paracetamol. The fragmentation of paracetamol was due to the elimination of acetyl fragment of the N-acetyl group to produce a primary 110 m/z ammonium ion. The specificity of the method to each analyte was achieved by matching the retention time of each analyte in the sample to that in the standard and by detecting the protonated molecular ion plus a specific fragment (152 or 110 m/z) for each analyte. Further confirmation was achieved by spiking the samples with the standard compounds, followed by monitoring the growth of each peak. The urine of saline-treated mice (blank) yielded no detectable peaks on the chromatograph, whereas urine
Selective determination of paracetamol metabolites / A.K. Hewavitharana et al. / Anal. Biochem. 374 (2008) 106–111
samples from paracetamol-treated mice showed peaks corresponding to paracetamol and its metabolites (Fig. 2). As seen in Fig. 2, both PG and PS produced two MRM peaks at the same retention time as the main peak: one due to the monitored MRM transition (328 fi 152 m/z ion for PG and 232 fi 152 m/z ion for PS) and another due to an additional secondary transition of 152 fi 110 m/z ion (for both PG and PS because both produce 152 m/z ion from the primary transition) . Because the method is set up to detect 152 fi 110 m/z ion (as the primary transition for paracetamol), the secondary transition of PG and PS (which is the same as the primary transition of paracetamol) is also detected at their respective retention times (second trace of PG and PS peaks in Fig. 2). This gave an additional degree of confirmation for PG and PS (in addition to retention time and the primary MRM transition), increasing the specificity of the method for these two analytes. However, for PS two separate urinary PS peaks were observed at different retention times (PS1 and PS2). Because the standard chromatogram contained only one peak for PS and the chromatogram of the blank urine showed no peaks, the additional peak most likely is the result of paracetamol metabolism in mice. The appearance of two peaks was not due to a chromatographic problem such as peak splitting (due to column overloading) because the PS in the standard solution did not appear as a split peak even at very high concentrations. Spiking confirmed that PS1 was the peak due to the standard compound
Fig. 2. Representative chromatogram of a mouse urine sample at 2 h after paracetamol administration (250 mg/kg).
109
PS, in which the sulfate group is in the para position to the N-acetyl group (shown as PS in Fig. 1). One way to explain the appearance of PS2 is that it may have been due to a secondary transition of a metabolite that was not detected by the method described here. The metabolites suspected were paracetamol cysteinate (PC, MW = 270) and paracetamol mercapturate (PM, MW = 312) (Fig. 1), with the suspected primary transitions 271 fi 232 m/z ion and 313 fi 232 m/z ion, respectively. If this is the case, then the secondary transition of 232 fi 152 m/z ion could have caused the PS2 peak. When the primary transitions 271 fi 232 m/z ion and 313 fi 232 m/z ion were monitored for urine samples, no peaks were observed, implying that the PS2 peak was not due to the secondary transition of either PC or PM. For a thorough investigation, a total ion chromatogram (TIC) was run for a sample within the range of 200 to 350 m/z. An attempt was made to extract a variety of suspected metabolite ions from TIC: 313 (PM), 271 (PC), and 232 (paracetamol-4-sulfate [N-acetyl4-aminophenol sulfate] or paracetamol-3-sulfate [N-acetyl3-aminophenol sulfate]). In addition to these, we attempted to extract three further putative metabolites from the TIC: paracetamol N-sulfate (N-acetyl-N-sulfonyl-4-aminophenol), 248 (N-acetyl-3-sulfonyl-4-aminophenol) (II), and 263 (N-acetyl-3-sulfonyl-4-aminomethoxyphenol) (III) (Fig. 1). The extracted ion chromatogram from TIC, however, showed only one peak that was PC. The other metabolites were not in sufficiently high concentrations to be detected in TIC mode. To further confirm that the PC is not responsible for producing the PS2 peak, a fragment ion (MS2) chromatogram was run for 271 (PC) ion in the sample. This produced its major fragment at 152 (same as PG and PS) rather than at 232. Furthermore, the MS2 chromatogram of the sample for 232 ion produced two peaks, at the retention times of PS1 and PS2, with identical fragment ion spectra (with 152 m/z as the major fragment). This chromatogram and spectra are shown in Fig. 3. In conclusion, the PS2 peak is produced by an isomer of PS that has the same molecular weight. MS is limited to provide information about the position of the sulfate group. Possible compounds are paracetamol-3-sulfate (i.e., sulfate group is in the meta position to the N-acetyl group) and paracetamol N-sulfate (I). Standard compounds are not commercially available for either of these compounds; therefore, we were unable to confirm the identity of PS2. In the literature, there was no mention of the formation of an additional isomer of PS during paracetamol metabolism, and we suspect this is due to the nonspecific nature of the detection (by UV) commonly used to monitor the metabolic products of paracetamol. Because there were many additional peaks (other than the peaks eluting at the retention times of the metabolite standards) with UV detection, an additional peak such as PS2 would have been ignored as a matrix peak. In this study, the use of tandem MS for the detection of paracetamol metabolites provides three degrees of specificity (retention time, MS1, and MS2) for each metabolite, whereas the UV
110
Selective determination of paracetamol metabolites / A.K. Hewavitharana et al. / Anal. Biochem. 374 (2008) 106–111
Fig. 3. Chromatogram of a mouse urine sample after paracetamol administration showing the fragmentation of 232 ion detected in MS2 mode. The fragmentation patterns of the PS1 and PS2 peaks are shown in the middle and lower panels, respectively.
method relies on only one degree of specificity (retention time). Notarianni and coworkers [17] detected a second peak for PS, and it was identified as one of the minor metabolites (paracetamol-3-sulfate). The order of elution is also consistent with our results where the para-substituted sulfate elutes before the meta substitute. This trend is also evident with the order of elution of paracetamol, which is the para substitute, and the IS, which is the meta substitute. We also found that the IS response in tandem MS is more sensitive than the paracetamol response. If the sulfates also follow the same trend, then it is likely that PS2, although detected as a significant peak, is likely to occur at very small concentrations compared with PS1. Taken together, the current and previous [17] studies suggest that the PS2 peak is most likely to be paracetamol-3sulfate. Quantitation of the metabolites in urine samples was carried out using paracetamol, PG, and PS standards and the structurally similar IS 3-acetamidophenol. Calibration curves consisted of five points for each analyte and were linear from 0 to 1.2 mmol/L for paracetamol (R2 = 0.998), from 0 to 0.6 mmol/L for PG (R2 = 0.996), and
from 0 to 0.06 mmol/L for PS (R2 = 0.969) with 30-ll injection volumes. The ranges of concentrations observed in mouse urine 2 h after intraperitoneal injection of 250 mg/ kg paracetamol (after correction for dilution) were 1.05 to 2.68 mmol/L for paracetamol, 1.74 to 2.48 mmol/L for PG, and 0.05 to 0.13 mmol/L for PS. The on-column limit of detection (LOD) determined, using standards prepared by spiking urine with each analyte and estimated as three times noise, were 2.4 pmol for paracetamol and 1.2 pmol for PG and PS. Considering the sample dilution step and the volume injected (30 ll), these equate to 0.66 lmol/L of paracetamol and 0.33 lmol/L of PG and PS in urine. These limits are lower than those reported previously [2] for UV detection: 2.2 lmol/L (for paracetamol), 0.52 lmol/L (for PG), and 1.4 lmol/L (for PS). The repeatability values (n = 8), determined by analyzing multiple samples of the same batch on the same day and expressed as percentage standard deviations, were 2.08% for paracetamol, 1.62% for PG, and 7.40% for PS. The method has high specificity due to tandem MS detection, and it is consistent and fast with a total run time of 30 min, including a column wash and a 10-min column reequilibration.
Selective determination of paracetamol metabolites / A.K. Hewavitharana et al. / Anal. Biochem. 374 (2008) 106–111
Conclusion The LC–MS/MS technique was applied to detect and quantify paracetamol metabolites in biological fluids. The unique attributes of this technique enabled the detection and quantitation of metabolites with high specificity and sensitivity. The method may be adapted for the determination of paracetamol metabolites in other biological fluids of human and other animals.
[7]
[8]
[9]
Acknowledgments This study was funded in part by NHMRC and ARC grants to D.M. S.L. is a recipient of the University of Queensland Confirmation Scholarship.
[10]
[11]
References [1] J.R. Mitchell, D.J. Jollow, S.W. Potter, D.C. Davis, J.R. Gillette, B.B. Brodie, Acetaminophen-induced hepatic necrosis: I. Role of drug metabolism, J. Pharmacol. Exp. Ther. 187 (1973) 185–194. [2] M.V. Vertzoni, H.A. Archontaki, P. Galanopoulou, Development and optimization of a reversed-phase high-performance liquid chromatographic method for the determination of acetaminophen and its major metabolites in rabbit plasma and urine after a toxic dose, J. Pharm. Biomed. Anal. 32 (2003) 487–493. [3] A. Di Girolamo, W.M. O’Neill, I.W. Wainer, A validated method for the determination of paracetamol and its glucuronide and sulphate metabolites in the urine of HIV+ /AIDS patients using wavelengthswitching UV detection, J. Pharm. Biomed. Anal. 17 (1998) 1191– 1197. [4] A.W. Abu-Qare, M.B. Abou-Donia, A validated HPLC method for the determination of pyridostigmine bromide, acetaminophen, acetylsalicylic acid, and caffeine in rat plasma and urine, J. Pharm. Biomed. Anal. 26 (2001) 939–947. [5] T. Matsumoto, T. Sano, T. Matsuoka, M. Aoki, Y. Maeno, M. Nagao, Simultaneous determination of carisoprodol and acetaminophen in an attempted suicide by liquid chromatography–mass spectrometry with positive electrospray ionization, J. Anal. Toxicol. 27 (2003) 118–122. [6] A. Panusa, G. Multari, G. Incarmato, L. Gagliardi, High-performance liquid chromatography analysis of anti-inflammatory pharmaceuticals with ultraviolet and electrospray–mass spectrometry
[12]
[13]
[14]
[15]
[16]
[17]
111
detection in suspected counterfeit homeopathic medicinal products, J. Pharm. Biomed. 43 (2007) 1221–1227. I.R. Miksa, M.R. Cummings, R.H. Poppenga, Multi-residue determination of anti-inflammatory analgesics in sera by liquid chromatography–mass spectrometry, J. Anal. Toxicol. 29 (2005) 95–104. S. La, H.H. Yoo, D.H. Kim, Liquid chromatography–mass spectrometric analysis of urinary metabolites and their pattern recognition for the prediction of drug-induced hepatotoxicity, Chem. Res. Toxicol. 18 (2005) 1887–1896. L.D. Betowski, W.A. Korfmacher, J.O. Lay Jr., D.W. Potter, J.A. Hinson, Direct analysis of rat bile for acetaminophen and two of its conjugated metabolites via thermospray liquid chromatography mass spectrometry, Biomed. Environ. Mass Spectrom. 14 (1987) 705–709. B.L. Ackermann, J.T. Watson, J.F. Newton Jr., J.B. Hook, W.E. Braselton Jr., Application of fast atom bombardment mass spectrometry to biological samples: Analysis of urinary metabolites of acetaminophen, Biomed. Mass Spectrom. 11 (1984) 502–511. T.A. Getek, W.A. Korfmacher, T.A. McRae, J.A. Hinson, Utility of solution electrochemistry mass spectrometry for investigating the formation and detection of biologically important conjugates of acetaminophen, J. Chromatogr. 474 (1989) 245–256. M. Ohta, N. Kawakami, S. Yamato, K. Shimada, Analysis of acetaminophen glucuronide conjugate accompanied by adduct ion production by liquid chromatography–atmospheric pressure chemical ionization–mass spectrometry, J. Pharm. Biomed. 30 (2003) 1759– 1764. O.Q.P. Yin, S.S.L. Lam, M.S.S. Chow, Simultaneous determination of paracetamol and dextropropoxyphene in human plasma by liquid chromatography/tandem mass spectrometry: Application to clinical bioequivalence studies, Rapid Commun. Mass Spectrom. 19 (2005) 767–774. X. Chen, J. Huang, Z. Kong, D. Zhong, Sensitive liquid chromatography–tandem mass spectrometry method for the simultaneous determination of paracetamol and guaifenesin in human plasma, J. Chromatogr. B 817 (2005) 263–269. C. Celma, J.A. Allue, J. Prunonosa, C. Peraire, R. Obach, Simultaneous determination of paracetamol and chlorpheniramine in human plasma by liquid chromatography–tandem mass spectrometry, J. Chromatogr. A 870 (2000) 77–86. S. Lee, P.A. Dawson, A.K. Hewavitharana, P.N. Shaw, D. Markovich, Disruption of NaS1 sulfate transport function in mice leads to enhanced acetaminophen-induced hepatotoxicity, Hepatology 43 (2006) 1241–1247. L.J. Notarianni, H.G. Oldham, P.N. Bennett, C.C.B. Southgate, R.T. Parfitt, Epoxides from paracetamol: A possible explanation for paracetamol toxicity, Adv. Exp. Med. Biol. 136 (1982) 1077–1083.