Journal of Chromatography B, 951–952 (2014) 7–15

Contents lists available at ScienceDirect

Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb

Simple, fast and sensitive LC–MS/MS analysis for the simultaneous quantification of nicotine and 10 of its major metabolites Markus Piller 1 , Gerhard Gilch 1 , Gerhard Scherer, Max Scherer ∗ ABF, Analytisch-Biologisches Forschungslabor München, Goethestrasse 20, 80336 Munich, Germany

a r t i c l e

i n f o

Article history: Received 10 October 2013 Accepted 10 January 2014 Available online 24 January 2014 Keywords: Biomarkers of exposure (BoE) LC–MS/MS Mass spectrometry Nicotine metabolism Urine

a b s t r a c t Urinary determination of nicotine metabolites provides an ideal tool for the quantitative assessment of the tobacco use-related nicotine dose, provided that the considered metabolites comprise a large share of the amount taken up. A method based on liquid chromatography–tandem mass spectrometry (LC–MS/MS) was developed for the sensitive, fast and robust analysis of nicotine and 10 major nicotine metabolites (“Nic+10”), including cotinine, trans-3 -hydroxy-cotinine, nicotine-N-glucuronide, cotinine-N-glucuronide, trans-3 -hydroxy-cotinine-O-glucuronide, nornicotine, norcotinine, nicotine-N oxide, cotinine-N -oxide and 4-hydroxy-(3-pyridyl)-butanoic acid. Corresponding deuterated internal standards were spiked prior to a simple and straightforward solid phase extraction (SPE) procedure. Liquid chromatography was performed on a reversed phase C8 column and mass-specific detection was conducted in scheduled-MRM mode. The method was validated according to FDA Guidelines, showing excellent selectivity, precision, accuracy and robustness. The limits of quantification were in the range 0.2–2.3 ng/ml for all analytes. The novel method was applied to human urine samples derived from 25 smoking subjects. Quantitative results were correlated against a previously used LC–MS/MS method and compared to reports from the literature. The relative molar profile of nicotine and its 10 major metabolites was in good agreement with the literature. In addition, correlation amongst the two methods was excellent for almost all analytes, whereas the accordance between both methods was moderate for hydroxy-cotinine-O-glucuronide and norcotinine. These deviations, however, could be explained. The current method allows the simultaneous determination of nicotine and its 10 major metabolites (metabolite coverage about 95% of the absorbed dose) from a small sample volume and within a reasonable amount of time. Due to its wide dynamic range, high sensitivity and high throughput capabilities, this method could serve as a powerful tool for quantifying the nicotine dose of smokers, passive smokers as well as novel tobacco and nicotine product users in clinical and epidemiological studies. © 2014 Elsevier B.V. All rights reserved.

1. Introduction It is well known that tobacco consumption is one of the most critical public health problems, making tobacco use the leading cause of premature death in developed countries [1–3]. Tobacco alkaloids are the active principal components in all tobacco products. Among more than 20 different alkaloids found in tobacco,

Abbreviations: COPD, chronic obstructive pulmonary disease; FDA, Food and Drug Administration; Nic, nicotine; Cot, cotinine; OH-Cot, trans-3 -hydroxycotinine; Nic-Gluc, nicotine-N-glucuronide; Cot-Gluc, cotinine-N-glucuronide; OH-Cot-Gluc, trans-3 -hydroxy-cotinine-O-glucuronide; NorNic, nornicotine; NorCot, norcotinine; NNO, nicotine-N -oxide; CNO, cotinine-N -oxide; HyPyBut, 4-hydroxy-(3pyridyl)-butanoic acid; UGT, UDP-glyucuronosyltransferase; FMO, flavin containing monooxygenase. ∗ Corresponding author. Tel.: +49 89 535395; fax: +49 89 5328039. E-mail address: [email protected] (M. Scherer). 1 Shared first authors. 1570-0232/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jchromb.2014.01.025

nicotine is the most abundant (98% of the total alkaloids) and accounts for widespread human use of tobacco products throughout the world, probably due to its pharmacological effects and possibly also its addictive potential [4–8]. Especially tobacco smoking is involved in a multitude of chronic diseases such as cancer (in particular lung cancer), cardiovascular diseases (CVD) and chronic obstructive pulmonary disease (COPD) [1]. Nicotine is rapidly and extensively metabolized by the liver in various compounds upon absorption in the human body [9,10]. Nicotine metabolism and relative amounts of the urinary metabolites are highlighted in Fig. 1. The predominant pathway during first pass metabolism is C-oxidation of nicotine to form cotinine. In humans about 70–80% of nicotine is converted to cotinine [11] which is subsequently hydroxylated, glucuronidated, oxidized and de-methylated to form various cotinine-derived metabolites (Fig. 1) [9]. N-Oxidation is also a primary route of nicotine metabolism, although only 4–7% of the nicotine amount absorbed by smokers is being transformed via this direction [9]. In addition to the oxidation of the pyrrolidine

8

M. Piller et al. / J. Chromatogr. B 951–952 (2014) 7–15

O N

+

N

N

O

N

O

+

N Glu

Nicotine-N-glucuronide (Nic-Glu;~2 %)

+

N

Nicotine-N`-oxide (NNO; ~7 %)

Glu

Cotinine-N-glucuronide (Cot-Glu; ~10 %)

N

N

N N

Nornicotine (NorNic; ~5 %)

+

Cotinine-N-Oxide (CNO; ~3%)

N

N

N N

Norcotinine (NorCot; ~2 %)

O

OH

Trans-3`-OH-cotinine (OH-Cot; ~30 %)

Cotinine (Cot; ~10 %)

Nicotine (Nic; ~5 %)

O

O

N

N

CO2H N OH 4-Hydroxy-4-(3-pyridyl)butanoic acid (HyPyBut; ~5-10 %)

N

N

O

N

N

O

O

Glu Trans-3`-OH-cotinine-glucuronide (OH-Cot-Gluc; ~15 %)

Fig. 1. Metabolism of nicotine to its major urinary metabolites. Abbreviations and relative amounts of metabolites in human urine are given in parenthesis, estimated from own and published data [9].

ring, nicotine is also de-methylated and glucuronidated to form the minor metabolites Nor-Nicotine and Nicotine-N-glucuronide, respectively. These metabolites account only for 5% respectively 2% of the nicotine dose absorbed [11]. About 10–15% of nicotine and metabolites is excreted as 4-oxo-4-(3-pyridyl)-butanoic acid and 4-hydroxy-4-(3-pyridyl)-butanoic acid in smokers urine (Fig. 1) [12]. For the first time in the year 2000, Hecht and colleagues have shown [12] that these metabolites arise directly from a nicotine transformation and not as initially thought from cotinine [13]. The determination of cotinine in blood, saliva and urine is well established and has been most frequently used as a biomarker of nicotine and tobacco smoke exposure [3,10]. Its suitability and limitations of this purpose has been extensively reviewed [14–18]. If a high percentage of the metabolites excreted in urine is covered, the determination of nicotine and its major metabolites provides a useful tool for estimating the total nicotine dose obtained by the various forms of tobacco use, which is difficult to quantify by other methods In contrast to measuring only single or a few metabolites, the high coverage rate of nicotine metabolites would also have the advantage that factors which affect the metabolism such as genetics, gender, enzymatic induction, inference with other chemicals would be of only minor importance. Various combinations of urinary nicotine metabolites have been applied measuring the nicotine uptake by smoking and other tobacco uses (for review, see [18]). Most frequently, nicotine and its major metabolites cotinine, trans-3 -hydroxy-cotinine, nicotine-Nglucuronide, cotinine-N-glucuronide, trans-3 -hydroxy-cotinineO-glucuronide, nornicotine, norcotinine, nicotine-N -oxide and cotinine-N -oxide have been determined in urine for this purpose. The chemical structures as well as approximate percentages of the nicotine uptake dose excreted in urine are shown in Fig. 1. The nicotine dose based on these 10 urinary metabolites (also called “Nic+9”) make up approximately 90% of the amount taken up. In three studies, “Nic+8” was determined by omitting nornicotine [19–21]. In other studies, the 10 mentioned metabolites without norcotinine [22] and cotinine-N -oxide [23] was determined. The proportions of the metabolites were found to be similar in the

various studies [18] (Fig. 1). “Nic+8” (nornicotine omitted) was also applied for nicotine dose measurements in two smoking behavior studies, however, without providing percentage for the single urinary metabolites [24,25]. Nicotine equivalents comprising “Nic+5” (nicotine, cotinine, OH-Cot and their respective glucuronides) represent approximately up to 80% of the nicotine [26]. [18]. “Nic+5” has been frequently applied as biomarker for the smoking- and tobacco userelated exposure to nicotine in recent years [27–34]. “Nic+5” equivalents in urine are determined by ‘indirect’ or ‘direct’ methods. Indirect methods comprise two analytical runs: (i) Determination of free bases (aglycons) in untreated urine and (ii) determination of the aglycons after enzymatic hydrolysis of the urine with ß-glucuronidase. The first determination provides values for the free bases (nicotine, cotinine, trans-3 hydroxycotinine). The second determination provides values for total bases (free + conjugated metabolite). The difference between the second and first determination represents the amount of the conjugated metabolite. Advantages of the indirect approach include the application of both gas chromatography (GC) and liquid chromatography (LC) with various detectors (NPD, MS, MS/MS), additionally, no unlabeled and labeled glucuronides are required as reference materials or internal standards, respectively. The disadvantage of the indirect approach is the fact that two separate analytical runs have to be performed and that the analytical variation for the glucuronides is, therefore, increased. This approach was used by various groups [22,26,35–37]. Direct methods [21,33,38] allow the determination of nicotine equivalents in urine in one run and do not require the enzymatic hydrolysis step. Disadvantages of this approach are requirement of relatively expensive analytical instrumentation as well as of unlabeled and labeled standards for the glucuronides. Of the three principal methodological approaches for the determination of nicotine metabolites in urine, namely LC, GC, immunological methods (for review, see [39]), LC–MS/MS is the most promising approach, because it allows the determination of the whole range of analytes, despite their largely deviating

M. Piller et al. / J. Chromatogr. B 951–952 (2014) 7–15

physico-chemical properties. In particular, it has been shown that also the glucuronide can be directly analyzed together with the aglycons and the other metabolites depicted in Fig. 1 [21,40]. McManus et al. [41] were able to determine nicotine and 17 of its metabolites in body fluids. However, this method did not include the glucuronides, which represent a quantitatively important fraction for nicotine dose assessment. The requirements for a suitable analytical method for quantifying the nicotine uptake in tobacco users are: (i) coverage of excreted metabolites should be close to 100%, (ii) high sensitivity, in order to quantify nicotine doses much lower than those obtained from conventional cigarettes, (iii) analysis in one run, allowing high throughput, (iv) high robustness. In order to meet these requirement, we decided to develop a LC–MS/MS method, which allows the simultaneous determination of the metabolites shown in Fig. 1 (“Nic+10”) accounting for more than 95% of the total nicotine dose absorbed. In addition to the “Nic+9” approach described above, 4-hydroxy-4-(3-pyridyl)butanoic acid (HyPyBut), representing about 5–10% of the nicotine dose, was also included in the method. In terms of sensitivity, the aim was to be able to quantify also the uptake of nicotine in nonsmokers exposed to environmental tobacco smoke (ETS). The novel method was validated according to FDA Guidelines [42] and applied to a series of smoker urine samples. Furthermore, the results were correlated against an independent LC–MS/MS method (Nic+8) previously developed in our laboratory [21], and compared to reports from the literature.

2. Materials and methods 2.1. Chemicals, standards, stock solutions and quality controls (−)Nicotine and (−)cotinine were purchased from Sigma (Deisenhofen, Germany). Nicotine-methyl-d3 was from CDN Isotopes (Quebec, Canada). (3 R, 5 S)-3 -hydroxycotinine-O-ˇd-glucuronide, (3 R, 5 S)-3 -hydroxycotinine-O-ˇ-d-glucuronide (N-CD3 ) were from Syntheselabor Dr. Mark (Worms, Germany). (3 R, 5 S)-3 -hydroxycotinine, transCotinine-methyl-d3 , 3 -hydroxycotinine methyl-d3 , nicotine-N-ˇ-glucuronide, nicotine-N-ˇ-glucuronide methyl-d3 , cotinine-N-ˇ-glucuronide, cotinine-N-ˇ-glucuronide methyl-d3 , 4-hydroxy-(3-pyridyl)butanoic acid, 4-hydroxy-(3-pyridyl)-butanoic acid d3 , (1 S, 2 S)-nicotine 1 -oxide, (1 S, 2 S)-nicotine 1 -oxid methyl-d3 , (S)-cotinine 1 -oxide, (S)-cotinine 1 oxide methyl-d3 , (R,S)nornicotine, (R,S)-nornicotine pyridyl-d4 , (R,S)-norcotinine and (R,S)-norcotinine pyridyl-d4 were obtained from Toronto Research Chemicals (Ontario, Canada). The purity of all reference compounds commercially available was ≥98%. Ammonium acetate (0.1%) in water was supplied by Fluka (Taufkirchen, Germany). Methanol (HPLC grade) was purchased from Promochem, (Wesel, Germany). Ultrapure water was prepared by a Seralpur Pro 90C apparatus (Seral, Minden, Germany). Formic acid (99.8%), ammonium hydroxide (99.8%) and HPLC grade solvents were purchased from Sigma–Aldrich (Taufkirchen, Germany). Oasis MCX cartridges (60 mg, 3 ml) were obtained from Waters Corp. (Milford, MA, USA). Stock solutions were prepared as aqueous solution at a concentration of 1 mg/ml. Primary stock solutions were stored at −20 ◦ C. The mixed working solutions for IS and calibrators were prepared from the primary stock solution in water. Working solutions were stored at −20 ◦ C until analysis. Under these conditions stock and working solutions were found to be stable for more than a year. A non-smoker urine-pool was spiked with 5 ng/ml of analyte concentration to generate the low level quality control (QC) sample.

9

Medium and high level QC samples were obtained from a smoking study. QC samples were aliquoted and stored at −20 ◦ C until use. 2.2. Calibration standards Quantification was performed with the standard addition method. A set of 10 calibrators were analyzed with each batch of unknown samples by spiking analyte-free non-smoker urine (no detectable amounts of analytes > LOQ) with increasing amounts of analytes, yielding a calibration range 1.5–1000 ng/ml. The same amount of IS-Mix was added to each calibration level (200 ng/ml urine). A weighted, 1/x (where x is the standard concentration), least-square model was fitted to the calibration line. 2.3. Data analysis Analyst software (version 1.5.2, Applied Biosystems, Foster City, CA, USA) was used for peak integration, calibration, and quantification. Processed data were transferred to Excel (Microsoft 2007, Redmond, USA) to perform further statistical analysis. 2.4. Study samples Study samples were obtained from 25 smokers (9–18 cigarettes/day). Subjects were healthy individuals participating in a diet controlled smoking study. The study was carried out under the German national guidelines and the protocol was approved by the ethic commission of the Medical Chamber of Nordrhein-Westfalen, Germany. Urine samples were collected at two different time points (spot urines) and after 24 h, however only the 24 h urine samples were considered for analysis. Urine samples were kept at −20 ◦ C until analysis. Aliquots of 50 ␮l were taken from each tube and used for analysis. 2.5. Sample preparation Sample preparation was performed as previously described by Ranagiah et al. [37] with minor modifications. Briefly, aliquots of 500 ␮l (low analyte levels) or 50 ␮l (high analyte levels, e.g. smoking subjects) of urine were thawed at room temperature and mixed with 10 ␮l of 10 ␮g/ml deuterated internal standard (IS) solution (Nic-d3 , Cot-d3 , OH-Cot-d3 , Nic-Gluc-d3 , OH-Cot-Gluc-d3 , HyPyBut-d3 , NorNic-d4 , NorCot-d4 , NNO-d3 and CNO-d3 in water). SPE columns (Oasis, MCX 60 mg, Waters, Milford, USA) were preconditioned with 3 ml each of methanol, water and ammonium formate buffer (20 mM, pH 2.5). The pH of urine samples was reduced to approximately 3 by adding 6 ␮l of 50% formic acid prior to loading on the SPE cartridges. SPE cartridges were then washed with 3 ml ammonium formate buffer (20 mM, pH 2.5). Analytes were eluted with 1 ml methanol/ammonium hydroxide (90/10; v/v) under strong basic conditions (pH 12). The eluate was evaporated in a SpeedVac (Jouan RC10.22, Thermo Scientific, Dreieich, Germany) to approximately 5–10 ␮l residual volume to avoid loss of volatile compounds such as nicotine and cotinine. Samples were further reconstituted in 100 ␮l water and 10 ␮l of the solution was injected into the LC–MS/MS system. 2.6. LC–MS/MS analysis Liquid chromatography was performed with an AT 1100 system including a binary pump, an autosampler and a column oven (Agilent Technologies, Waldbronn, Germany). A hybrid triple quadrupole mass spectrometer API 4000 equipped with a Turbo V source ion spray, operating in positive ESI mode, was used for detection (AB Sciex, Darmstadt, Germany). High purity nitrogen

10

M. Piller et al. / J. Chromatogr. B 951–952 (2014) 7–15

Table 1 MS parameters (Quan = quantifier, Qual = qualifier, IS = internal standard, CE = collision energy, RT = retention time) and limits of detection (LOD) and quantification (LOQ) for nicotine and its 10 major metabolites (“Nic+10”). Analyte

MRM [m/z]

IS (MRM) [m/z]

CE [V]

RT [min]

LOD (LOQ) [ng/ml]

Nic

162.9 → 130.0 (Quan) 162.9 → 132.0 (Qual) 177.2 → 80.1 (Quan) 177.2 → 98.0 (Qual) 193.0 → 80.1 (Quan) 193.0 → 134.1 (Qual) 339.2 → 162.9 (Quan) 353.1 → 177.0 (Quan) 369.1 → 118.1 (Quan) 149.1 → 131.9 (Quan) 149.1 → 129.8 (Qual) 163.1 → 79.9 (Quan) 163.1 → 84.0 (Qual) 179.1 → 129.9 (Quan) 179.1 → 117.0 (Qual) 193.1 → 96.1 (Quan) 193.1 → 79.1 (Qual) 182.1 → 108.1 (Quan)

Nic-d3 (166.0 → 130.2)

23 29 41 31 39 27 23 25 49 17 23 35 35 33 39 31 51 49

8.5

0.50 (1.50)

7.8

0.17 (0.51)

7.2

0.21 (0.63)

6.2 2.6 6.3 7.0

0.17 (0.51) 0.33 (0.99) 0.75 (2.25) 0.40 (1.20)

7.5

0.21 (0.63)

6.6

0.07 (0.21)

6.1

0.78 (2.34)

5.6

0.21 (0.63)

Cot OH-Cot Nic-Gluc Cot-Gluc OH-Cot-Gluc NorNic NorCot NNO CNO HyPyBut

Cot-d3 (180.1 → 80.0) OH-Cot-d3 (195.9 → 80.1) Nic-Gluc-d3 (342.2 → 166.2) Cot-Gluc-d3 (356.1 → 180.1) OH-Cot-Gluc-d3 (372.2 → 196.1) NorNic-d3 (153.1 → 134.0) NotCot-d3 (167.2 → 84.0) NNO-d3 (182.1 → 130.0) CNO-d3 (196.2 → 96.0) HyPyBut-d3 (184.9 → 108.9)

was produced by a nitrogen generator NGM 22-LC/MS (cmc Instruments, Eschborn, Germany). Chromatographic separation was achieved by gradient elution on a Luna 3 ␮m C8, 4.6 mm × 150 mm column (Phenomenex, Aschaffenburg, Germany) with 0.1% ammonium acetate (pH 7.0, solvent A) and MeOH (solvent B) as mobile phases. The column oven was maintained at 50 ◦ C. Chromatographic separation was optimized to obtain good resolution amongst the “Nic+10” metabolites. Gradient elution started with 95% A for 1 min, followed by a linear decrease to 20% A until 6 min, a step to 0% A within 0.1 min, 0% A until 8 min and re-equilibration at 95% A from 8.1 until 11 min. The flow rate was set to 700 ␮l/min throughout the entire run. The turbo ion spray source was operated in the positive ionization mode using the following settings: ion spray voltage = 5500 V, ion source heater temperature = 600 ◦ C, source gas 1 = 30 psi, source gas 2 = 40 psi, and curtain gas setting = 30 psi. Since numerous mass transitions are required to cover the 11 metabolites and their respective IS, analytes were monitored in scheduled multiple reaction monitoring (SMRM) mode, applying a retention time window of 90 s. Target scan time was set to 0.25 s. Quantifier and qualifier (as far as available) mass transitions, MS parameters and retention times (RT) are shown in Table 1. Quadrupoles Q1 and Q3 were working at unit resolution. 2.7. Method validation Urine samples for validation experiments were either obtained by spiking analyte-free non-smoker urine or pooled samples from smoker urine. Validation procedure was performed according to FDA Guidelines [42]. 2.7.1. Specificity The specificity of the method was analyzed by comparing retention times of analyte-free matrix with reference materials in blank samples. For each compound 2 MRMs (as far as available) were monitored to ensure that no interference at the analyte RT is present. 2.7.2. Precision In order to assess the precision of the analytical workflow, urine samples at three different concentration levels (low, medium and high) were analyzed 6 times in a row and on 6 consecutive days to determine intra-assay and inter-assay variations, respectively. The low concentration level was prepared by spiking non-smoker urine-pool with 5 ng/ml of analytes. Medium and high

concentration levels were taken from smoker urine-pools at the middle and the upper end of the calibration range. 2.7.3. Accuracy Assay accuracy was calculated using non-smoker urine samples spiked at 3 different concentration levels (5 ng/ml, 100 ng/ml and 1000 ng/ml), covering the entire calibration range. Each level was analyzed 5 times in a row. 2.7.4. LOD/LOQ The limit of detection (LOD) in urine, defined as a signal-tonoise-ratio (S/N) of 3, for nicotine and the ten nicotine-derived metabolites was determined by spiking three analyte-free matrix samples with a low concentration of analyte (S/N max. 30). The limit of quantification (LOQ) was calculated as 3 times the LOD. 2.7.5. Recovery Extraction efficiencies were determined by spiking analyte-free urine samples at 3 different concentration levels (low, medium, high) prior and after extraction. Samples were analyzed as triplicates. 2.7.6. Matrix effect The matrix effect was evaluated by comparing urine samples spiked with low and high analyte concentration post-extraction with reference standards at the same concentration in water. Each set of samples was analyzed in triplicates. 2.7.7. Carryover A high level smoker urine pool was injected 5 times in a row followed by the analysis of 3 non-smoker urines containing the analytes at the LOQ. This procedure was repeated 3 times. There would be no carryover if the CV of the low samples at the LOQ differed by 80% for all analytes and did not significantly vary between the 3 concentration levels tested (CV < 10%). Matrix effects were quantitatively assessed by analyzing standard compounds in water and spiked at the same amount into non-smoker urine matrix after extraction. We found a decline in signal intensity due to matrix interferences between 0% and 50%, depending on the analyte and matrix investigated. However, we could demonstrate that ion suppression could be fully compensated by using corresponding deuterated ISs (signal recovery A/IS of >90%). Matrix effects did not vary for the different concentration levels and were stable across the replicates within each series of analysis. No significant carryover was observed from high analyte containing urine samples into non-smoker urine samples spiked at the LOQ (CV < 18%). Calibration pre-tests showed that for the vast majority of analytes the linear dynamic range was already exceeded above 1000 ng/ml. In order to cope with the wide concentration range for nicotine and its metabolites (10 ␮g/ml for heavy smokers), we decided to rather prove accuracy after sample dilution than compromising the linear fit by a squared fit. Accuracies after sample dilution by a factor of 5, 10 and 50 were in the range between 89 and 113%. We characterized the stability of analytes in urine stored at different temperature and freeze–thaw cycles. All analytes were found to be stable stored for 24 h at room temperature. Similarly, post-preparative stability of the extracted urine samples was proven for 24 h at auto-sampler temperature (10 ◦ C). We also tested freeze–thaw stability of analytes in urine samples and found that all analytes were stable through 6 freeze–thaw cycles. 3.3. Method application The described LC–MS/MS method was applied to urine samples of 25 smokers from a diet controlled study. The mean urinary concentrations and the relative distribution for nicotine and its 10 major metabolites are summarized in Suppl. Table 1. For comparison, the relative amounts of the individual nicotine metabolites in smokers urine as reported in 7 studies from the literature

10.2% ± 5.6% Piller et al. (this study, N = 25)

5.6% 2.2%

The mean molar percentages are compared with 7 reports from the literature. Nic, nicotine; Cot, cotinine; OH-Cot, trans-3 -hydroxycotinine; Nic-Gluc, nicotine-N-glucuronide; Cot-Gluc, cotinine-N-glucuronide, OH-Cot-Gluc, trans-3 -hydroxycotinine-O,N-glucuronide; NorNic, nornicotine; NorCot, norcotinine; NNO, nicotine-N -oxide; CNO, cotinine-N-oxide; HyPyBut, 4-hydroxy-(3-pyridyl)-butanoic acid; n.d., not determined.

HyPyBut

5.0% ± 0.8% 2.8% ± 0.5% 4.4% ± 2.0% 1.3% ± 0.2% 0.7% ± 0.2% 14.3% ± 5.5% 16.7% ± 4.5% 5.0% ± 2.4% 25.2% ± 8.2%

5.1% 8.4% 5.5% 7.2% 2.9% 1.6% 8.8% 14.9% 5.6% 4.2%

14.2% ± 3.3%

0.9% ± 0.9% 6.8% ± 2.9% 3.7% ± 0.9% n.d. 6.7% 4.2% ± 1.8% 4.5% ± 2.0%

CNO NNO

3.0% ± 2.1% 6.8% ± 2.9% 3.7% ± 0.9% n.d. 6.7% 4.2% ± 1.8% 4.5% ± 2.0% n.d. 1.5% ± 0.5% n.d. n.d. 0.6% 2.1% ± 0.4% 2.2% ± 0.5%

NorCot NorNic

– – 0.6% ± 0.2% – – 0.5% ± 0.3% 6.9% ± 7.1% 10.0% 3.8% 5.9% 7.6%

OH-Cot-Gluc

22.8% ± 8.5% ± 7.8% ± 10.3% ± 7.4% 6.8% ± 9.7% ± 5.4% 6.3% 7.8% 6.0%

Cot-Gluc

14.0% ± 17.5% ± 15.8% ± 12.1% ± 20.1% 14.8% ± 7.3% ± 2.5% 2.2% 2.9% 2.1%

4.5% ± 2.8% ± 4.6% ± 2.6% ± 3.7% 5.4% ± 1.3% ±

Nic-Gluc

10.6% 7.4% 12.5% 12.8% OH-Cot

9.2% ± 13.2% ± 13.3% ± 14.8% ± 15.2% 11.1% ± 8.6% ±

2.6% 3.9% 3.1% 5.9% Cot

9.4% ± 10.4% ± 10.4% ± 7.9% ± 9.5% 8.8% ± 7.7% ± [19] (N = 91) [20] (N = 11) [22] (N = 12) [36] (N = 4) [21] (N = 5) [23] (N = 166) [37] (N = 61)

5.7% 3.7% 4.4% 4.6% Nic Ref.

Nicotine is metabolized primarily by the liver enzymes CYP2A6, UDP-glyucuronosyltransferase (UGT) and flavin containing monooxygenase (FMO) [9,45,46]. Urinary nicotine and metabolite concentrations have been widely used to determine nicotine equivalents as an estimate of the nicotine dose that is taken up by smoking or use of other tobacco and nicotine products [37,28,32,47]. A valid estimation of the uptake, reflected by the urinary metabolite levels, highly depends on the accurate measurement. In order to obtain reliable data on the nicotine uptake and the subsequent metabolization, a high percentage of the dose taken up in the form of the major urinary metabolites should be assessed. In our study with smokers, we found a mean share of 5% for HyPyBut for the 11 metabolites measured (Table 3). Thus the newly developed “Nic+10” method reflects approximately 95% of the absorbed nicotine dose, which is an excellent rate for the intended purpose. Our intention was to develop a multi-analyte high-throughput LC–MS/MS method for the simultaneous quantification of nicotine and its 10 major metabolites from urine samples. The main goal during method development was to optimize and accelerate analyte quantification, thus making this method suitable for the application in larger clinical studies. We were able to improve the chromatographic efficiency, metabolite coverage, sensitivity and throughput relative to previous methods [21,35,37,38,44]. As shown in Fig. 2, the LC method resolves isobaric analyte pairs (e.g. Nic/NorCot and OH-Cot/CNO) which cannot be unambiguously specified by their mass transitions. By coupling selective chromatography with selective mass spectrometry of 28 separate MRMs, we achieved unambiguous quantification of the 11 target analytes within a single run. Table 4 summarizes the method performance against reports from the literature. As compared to similar multi-analyte methods for the simultaneous determination of nicotine and some of its metabolites [35,37,38], we achieved excellent LODs across the entire range of analytes, considering the relative wide range of polarity and small sample volume required (Table 4).

Table 3 Studies from the literature (values are mean molar percentages ± SD, table adapted and modified from Tricker 2006 [32]).

4. Discussion

36.1% ± 35.2% ± 39.1% ± 42.4% ± 34.1% 31.5% ± 26.8% ±

[35,37,38,43,44] and partly reviewed by Tricker [18] are shown in Table 3. As demonstrated in Table 3, results obtained with the current method are in good agreement with those reported in the literature. Supplementary material related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jchromb. 2014.01.025. For cross-validation purposes and in order to strengthen the validity of our established method, results were compared to those obtained with a previous LC–MS/MS method for the analysis of nicotine and its 8 major metabolites [21]. Correlation plots for all analytes are shown in Suppl. Fig. 1. Results obtained with both methods showed an excellent correlation for Cot, Nic, OH-Cot, Cot-Gluc and Nic-Gluc (R2 > 0.97). Correlation coefficients for OHCot-Gluc, NorCot and NNO were slightly lower (R2 : 0.88–0.89), however still within an acceptable range. Note that NorNic and HyPyBut were not included in the previous method. Supplementary material related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jchromb. 2014.01.025. Accuracies amongst the two methods, reflected by the slope of the linear regression, were in good agreement for Cot, Nic, OH-Cot, Cot-Gluc, Nic-Gluc, NNO and CNO, respectively (slope: 0.994–1.10). However, the new method revealed an underestimation by 50% for NorCot and an overestimation for OH-Cot-Gluc by 40% as compared to the previous method.

13

n.d. n.d. n.d. n.d. n.d. n.d. 9.7% ± 5.4%

M. Piller et al. / J. Chromatogr. B 951–952 (2014) 7–15

0.2 –2.3 Nic+10 (>95%) SPE (Oasis MCX) 0.05 (smokers) 0.5 (non-smokers) 11 Reversed-phase C8 Piller et al.

ESI positive

Nic+14 (≈99%) Nic+5 (60%–70%) (no glucuronides!) Nic+5 (≈70%) Nic+8 (≈70%) SPE (Oasis MCX) SPE (Oasis HLB) SPE (Oasis HLB) SPE (Oasis HLB and MCX) 0.5 1 0.5 1 35 15 Not reported 15 Pentfluorphenyl Pentafluorphenyl Diol Pentafluorphenyl Rangiah et al. [37] Xu et al. [44] Haevner et al. [35] Miller et al. [38]

ESI positive ESI positive APCI positive ESI positive

Nicotine metabolite (N) coverage (%) Ionization Urine volume [ml]

Sample clean-up Analysis time [min] LC type Authors (Ref.)

Table 4 Comparison of method performances.

300–600 0.1–1 6–30 1–50

M. Piller et al. / J. Chromatogr. B 951–952 (2014) 7–15

LOQ (ng/ml)

14

We performed a linear regression of the peak-area ratios versus concentrations with a 1/x weighting. Although the LC–MS/MS method provided excellent linearity over almost three orders of magnitude (1 ng/ml–1000 ng/ml), we observed detector signal saturation above 1000 ng/ml for most analytes. In order to cope with the wide dynamic range required to allow its application from heavy smokers (levels up to 10 ␮g/ml for some metabolites) to non-tobacco users (levels < 5 ng/ml) we tested the accuracy after sample dilution. We could demonstrate that the machine-derived limitation of the dynamic range can be circumvented by diluting the urine samples up to 50-fold prior to extraction. Hence, the improved dynamic range upon dilution along with the high sensitivity for this method allow its application to urine samples from heavy smokers to people using various tobacco and nicotine products to people supposedly exposed to environmental tobacco smoke (passive smokers). Moreover, our LC–MS/MS was validated according to FDA Guidelines [42]. We achieved excellent accuracy and precision rates (Table 2). Extraction yields (recoveries) ranged from 80% to 100% and ion suppression by matrix components amounted to 0–50%. However, we could show that eventual losses during SPE and variations in the ionization efficiency caused by matrix interferences were fully compensated by the corresponding deuterated ISs. Finally, we applied the newly developed method to urine samples derived from a nutrition-controlled smoking study. Urine from 25 smoking subjects (cigarette consumption: 9–18 cig/day) was analyzed for Nic and its major 10 metabolites. To further evaluate validity of our method, study results for smoking individuals were cross-validated against those obtained with a LC–MS/MS method previously developed in our laboratory [21]. As shown in Suppl. Fig. 1 both methods showed excellent correlations. Deviations between both methods were less than 10%, except for NorCot (−50% for the novel method) and OH-Cot-Gluc (+40%). The relatively large discrepancy for OH-Cot-Gluc was caused only by two urine samples containing the highest analyte concentrations. These concentration levels exceeded the linear dynamic range of the “old” method (10–5000 ng/ml) [21]. Detector signal saturation and, therefore, underestimation of the real concentration of these two samples by the “old” method might be a possible explanation. The reason for the overestimation by 50% of NorCot with the “old” method (which included no SPE purification step) could be explained by potentially interfering matrix components eluting at the same RT. Considering the validation data, we suggest that the LC–MS/MS method described here is the superior assay. In a second step we compared the relative molar concentrations with reported values from the literature (Table 4). Although only a relatively small number of subjects has been analyzed as “show case” application (n = 25), the mean relative molar concentrations of the determined Nic+10 metabolites are consistent with reported concentrations in smoker’s urine [35,37,38,43,44].

5. Conclusion In summary, we have developed a simple, fast and robust method for the simultaneous quantification of urinary nicotine and its 10 major metabolites, reflecting approximately 95% of the amount of nicotine absorbed. A coverage of 95% is of paramount importance for the reliable estimation of the nicotine dose, since analyzing only single or a subset of nicotine metabolites might be extensively influenced by inter-individual variations in the nicotine metabolism. Due to its wide dynamic range, high sensitivity and high throughput capabilities, this method could serve as a powerful tool to investigate nicotine uptake in clinical studies including diverse groups of subjects such as smokers, people using novel

M. Piller et al. / J. Chromatogr. B 951–952 (2014) 7–15

tobacco and nicotine products or people being exposed to low nicotine doses such as from environmental tobacco smoke. References [1] U.S. Department of Health and Human Services, How Tobacco Smoke Causes Disease: The Biology and Behavioral Basis for Smoking-Attributable Disease – A Report of the Surgeon General, National Library of Medicine Cataloging, Rockville, MD, USA, 2010. [2] International Agency for Research on Cancer, IARC Monogr. Eval. Carcinog. Risks Hum. 83 (2004) 1–1438. [3] S.S. Hecht, Nat. Rev. Cancer 3 (2003) 733–744. [4] J.W. Gorrod, P. Jenner, in: W.J. Hayes Jr. (Ed.), The Metabolism of Tobacco Alkaloids Essays in Toxicology, Academic Press, New York, London, 1975, pp. 35–78. [5] T. Kisaki, S. Maeda, A. Koiwai, Y. Mikami, T. Sasaki, H. Matsushita, Beitr. Tabakforsch. Int. 9 (1978) 308–316. [6] N.L. Benowitz, Annu. Rev. Pharmacol. Toxicol. 57 (2009) 57–71. [7] J.A. Dani, D.J. Balfour, Trends Neurosci. 34 (2011) 383–392. [8] U.S. Department of Health and Human Services, The Health Consequences of Smoking: Nicotine Addiction. A Report of the Surgeon General, Public Health Service, Rockville, MD, 1988. [9] J. Hukkanen, P. Jacob III, N.L. Benowitz, Pharmacol. Rev. 57 (2005) 79–115. [10] N.L. Benowitz, J. Hukkanen, P. Jacob III, Handb. Exp. Pharmacol. 192 (2009) 29–60. [11] N.L. Benowitz, P. Jacob Iii, Clin. Pharmacol. Ther. 56 (1994) 483–493. [12] S.S. Hecht, J.B. Hochalter, P.W. Villalta, S.E. Murphy, Proc. Nat. Acad. Sci. U.S.A. 97 (2000) 12493–12497. [13] H. McKennis, S.L. Schwartz Jr., E.R. Bowman, J. Biol. Chem. 239 (1964) 3990–3996. [14] N.L. Benowitz, Epidemiol. Rev. 18 (1996) 188–204. [15] R.A. Etzel, Prev. Med. 19 (1990) 190–197. [16] P.N. Lee, in: J.W. Gorrod, I.P. Jacob (Eds.), Uses and Abuses of Cotinine as a Marker of Tobacco Smoke Exposure Nicotine and Related Compounds and Their Metabolites, Elsevier, Amsterdam, 1999, pp. 669–719. [17] O. West, P. Hajek, H. McRobbie, Psychopharmacology (Berl.) 218 (2011) 313–322. [18] A.R. Tricker, Beitr. Tabakforsch. Int. 22 (2006) 147–175. [19] G. Andersson, E.K. Vala, M. Curvall, J. Oral Pathol. Med. 26 (1997) 117–123. [20] G.D. Byrd, K.M. Chang, J.M. Greene, J.D. deBethizy, Drug Metab. Dispos. 20 (1992) 192–197. [21] M. Meger, I. Meger-Kossien, A. Schuler-Metz, D. Janket, G. Scherer, J. Chromatogr, B. Analyt, Technol. Biomed. Life. Sci. 778 (2002) 251–261. [22] N.L. Benowitz, P. Jacob Iii, I. Fong, S. Gupta, J. Pharmacol. Exp. Ther. 268 (1994) 296–303. [23] W. Luck, H.J. Nau, Pediatrics 107 (1985) 816–820. [24] G.D. Byrd, R.A. Davis, W.S. Caldwell, J.H. Robinson, J.D. deBethizy, Psychopharmacology (Berl.) 139 (1998) 291–299.

15

[25] G.D. Byrd, J.H. Robinson, W.S. Caldwell, J.D. deBethizy, Psychopharmacology (Berl.) 122 (1995) 95–103. [26] M. Curvall, E. Kazemi-Vala, G. Englund, in: F. Adlkofer, K. Thurau (Eds.), Conjugation Pathway Effects of Nicotine on Biological Systems, Birkhäuser Verlag, Basel, 1991, pp. 69–75. [27] S. Feng, H.J. Roethig, Q. Liang, D. Oliveri, G. Scherer, 35th Annual Meeting of the American College of Clinical Pharmacology, Cambridge, Massachusetts, September 11, 2006, pp. 28–52. [28] P. Mendes, Q. Liang, K. Frost-Pineda, S. Munjal, R.A. Walk, H.J. Roethig, Regul. Toxicol. Pharmacol. 55 (2009) 17–27. [29] A. Morin, C.J. Shepperd, A.C. Eldridge, N. Poirier, R. Voisine, Regul. Toxicol. Pharmacol. (2010), http://dx.doi.org/10.1016/j.yrtph.2010.09.020. [30] R.S. Muhammad-Kah, A.D. Hayden, Q. Liang, K. Frost-Pineda, M. Sarkar, Regul. Toxicol. Pharmacol. 60 (2011) 79–83. [31] H.J. Roethig, S. Feng, Q. Liang, J. Liu, W.A. Rees, B.K. Zedler, J. Clin. Pharmacol. 48 (2008) 580–591. [32] H.J. Roethig, S. Munjal, S. Feng, Q. Liang, M. Sarkar, R.A. Walk, P.E. Mendes, Nicotine Tob. Res. 11 (2009) 1216–1225. [33] G. Scherer, J. Engl, M. Urban, G. Gilch, D. Janket, K. Riedel, Regul. Toxicol. Pharmacol. 47 (2007) 171–183. [34] C.J. Shepperd, A.C. Eldridge, D.C. Mariner, M. McEwan, M. Dixon, Regul. Toxicol. Pharmacol. 55 (2009) 97–109. [35] D.L. Heavner, J.D. Richardson, W.T. Morgan, M.W. Ogden, Biomed. Chromatogr. 19 (2005) 312–328. [36] S.S. Hecht, S.G. Carmella, S.E. Murphy, Cancer Epidemiol. Biomarkers Prev. 8 (1999) 907–913. [37] K. Rangiah, W.T. Hwang, C. Mesaros, A.I. Vachani, A. Blair, Bioanalysis 3 (2011) 745–761. [38] E.I. Miller, H.R. Norris, D.E. Rollins, S.T. Tiffany, D.G. Wilkins, J. Chromatogr. B: Analyt. Technol. Biomed. Life Sci. 878 (2010) 725–737. [39] R.A. Davis, M. Curvall, in: J.W. Gorrod, P. Jacob III (Eds.), Determination of Nicotine and Its Metabolites in Biological Fluids: In Vivo Studies Analytical Determination of Nicotine and Related Compounds and Their Metabolites, Elsevier, Amsterdam, 1999, pp. 583–643. [40] G.D. Byrd, M.S. Uhrig, J.D. deBethizy, W.S. Caldwell, P.A. Crooks, A. Ravard, R.M. Riggs, Biol. Mass Spectrom. 23 (1994) 103–107. [41] K.T. McManus, J.D. de Bethizy, D.A. Gareiz, G.A. Kyerematen, E.S. Vesell, J. Chromatogr. Sci. 28 (1990) 510–516. [42] Food and Drug Administration (FDA), U.S. Department of Health and Human Services, 2001, pp. 1–22, http://www.fda.gov/cderguidance/4252fnl.pdf (accessed 04.17.09). [43] H. Kataoka, R. Inoue, K. Yagi, K. Saito, J. Pharm. Biomed. Anal. 49 (2009) 108–114. [44] X. Xu, M.M. Iba, C.P. Weisel, Clin. Chem. 50 (2004) 2323–2330. [45] P. Jacob Iii, N.L. Benowitz, A.T. Shulgin, Pharmacol. Biochem. Behav. 30 (1988) 249–253. [46] D.B. Kandel, M.C. Hu, C. Schaffran, J.R. Udry, N.L. Benowitz, Am. J. Epidemiol. 165 (2007) 901–910. [47] A. Florescu, R. Ferrence, T. Einarson, P. Selby, O. Soldin, G. Koren, Ther. Drug Monit. 31 (2009) 14–30.

MS analysis for the simultaneous quantification of nicotine and 10 of its major metabolites.

Urinary determination of nicotine metabolites provides an ideal tool for the quantitative assessment of the tobacco use-related nicotine dose, provide...
677KB Sizes 2 Downloads 0 Views