G Model
ARTICLE IN PRESS
CHROMA-355364; No. of Pages 8
Journal of Chromatography A, xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Determination of bisphenol A, triclosan and their metabolites in human urine using isotope-dilution liquid chromatography–tandem mass spectrometry Gilles Provencher a , René Bérubé a , Pierre Dumas a , Jean-Franc¸ois Bienvenu a , Éric Gaudreau a , Patrick Bélanger a , Pierre Ayotte a,b,c,∗ a
Centre de toxicologie du Québec, Institut national de santé publique du Québec (INSPQ), 945 Wolfe, Québec, QC, Canada G1V 5B3 Axe Santé des populations et pratiques optimales en santé, Centre de recherche du CHU de Québec, 2875 boul. Laurier, Édifice Delta 2, Bureau 600, Québec, QC, Canada G1V 2M2 c Département de médecine sociale et préventive, Université Laval, Pavillon Ferdinand-Vandry, Québec, QC, Canada G1V 0A6 b
a r t i c l e
i n f o
Article history: Received 30 December 2013 Received in revised form 15 April 2014 Accepted 20 April 2014 Available online xxx Keywords: Bisphenol A Triclosan Glucuronide metabolites Sulfate metabolites Isotope dilution LC–MS/MS Human urine
a b s t r a c t Bisphenol A (BPA) and triclosan (TCS) are ubiquitous environmental phenols exhibiting endocrine disrupting activities that may be involved in various health disorders in humans. There is a need to measure separately free forms and conjugated metabolites because only the former are biologically active. We have developed sensitive methods using isotope-dilution liquid chromatography–tandem mass spectrometry for individual measurements of free BPA and TCS as well as their metabolites, BPA glucuronide (BPAG), BPA monosulfate (BPAS), BPA disulfate (BPADS), TCS glucuronide (TCSG) and TCS sulfate (TCSS) in urine. Comparative analyses of urine samples from 46 volunteers living in the Quebec City area using the new methods and a GC–MS/MS method previously used in our laboratory revealed very strong correlations for total BPA (Spearman’s rs = 0.862, p < 0.0001) and total TCS concentrations (rs = 0.942, p < 0.0001). Glucuronide metabolites were the most abundant BPA and TCS species in urine samples (>94% of total urinary concentrations). Unconjugated TCS concentrations represented a small proportion of total TCS species (median = 1.6%) but its concentration was likely underestimated due to losses by adsorption to the surface of polypropylene tubes used for sample storage. To our knowledge, we are the first to report levels of free, sulfated and glucuronidated TCS levels in human urine. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Bisphenol A (BPA) and triclosan (TCS) are two widely used phenolic compounds with numerous industrial and commercial applications. BPA is the monomer in the production of polycarbonate plastics and is also a component of epoxy resins that are used to produce food and beverage packaging materials and dental sealants among others [1]. Trace amounts of BPA have been shown to leach from polycarbonate containers [2–4], which led to the ban of polycarbonate baby bottles in Canada in 2010. BPA has been found in foodstuffs [5,6], and diet is most likely the major source of exposure to this compound [1,7,8]. TCS is a bactericidal agent added to soap and toothpaste [9,10]. Dermal absorption and ingestion of TCS in
∗ Corresponding author at: Centre de toxicologie du Québec, Institut national de santé publique du Québec (INSPQ), 945 Wolfe, Québec, QC, Canada G1V 5B3. Tel.: +1 418 650 5115x4654; fax: +1 418 654 2148. E-mail address: [email protected] (P. Ayotte).
these personal care products are thought to contribute the most to human exposure [9]. Both BPA and TCS are endocrine disrupting chemicals: BPA is an estrogen receptor agonist and an androgen receptor antagonist [11,12]. TCS interacts with constitutive androstane and pregnane-X receptors [13] and exhibits antagonistic activity in both estrogen receptor- and androgen receptor-responsive bioassays [14]. Human exposure to these compounds have been associated with various adverse health effects including reproductive and endocrine disorders, cardiovascular disease, obesity and metabolic syndrome, cancer, immune system dysfunction and neurobehavioral defects [1,15–17]. However, whether or not these associations indicate an important public health risk is highly controversial. An argument raised by those supporting a trivial health risk from exposure to these phenolic compounds is their rapid biotransformation to biologically-inactive metabolites [18]. Indeed, following their absorption by the gastro-intestinal tract, free BPA and TCS are rapidly and efficiently conjugated by hepatic uridine diphosphate-glucuronyltransferases to their glucuronidated forms
http://dx.doi.org/10.1016/j.chroma.2014.04.072 0021-9673/© 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: G. Provencher, et al., J. Chromatogr. A (2014), http://dx.doi.org/10.1016/j.chroma.2014.04.072
G Model CHROMA-355364; No. of Pages 8 2
ARTICLE IN PRESS G. Provencher et al. / J. Chromatogr. A xxx (2014) xxx–xxx
Fig. 1. Chemical structures of native and isotope-labeled compounds analyzed in this study. Internal standards for free BPA (13 C12 -BPA) and free TCS (13 C12 -TCS) bear 13 C atoms at all positions of aromatic rings.
and eliminated in urine [19,20]. However, the recent demonstration in the dog model of an efficient and rapid absorption of BPA through the oral mucosa by the sublingual route (avoiding the first-pass hepatic metabolism) could indicate a greater systemic exposure to free BPA by the oral route than previously thought [21]. Also, BPAG can be deconjugated by -glucuronidase found in placenta, liver, kidney, and intestine [18]. The development of an analytical method allowing the determination of unconjugated (“free”) and conjugated forms of BPA and TCS is central to an improved assessment of the health risk associated with exposure to these ubiquitous compounds. Both liquid chromatography–tandem mass spectrometry (LC–MS/MS) and gas chromatography–mass spectrometry (GC–MS) techniques have been used for the measurement of BPA and its metabolites in biological specimens [22–24]. Because standards of BPA conjugates were not commercially available, most methods measured only total BPA and TCS concentrations following enzymatic hydrolysis of conjugates [25–29]. Some researchers measured both free BPA concentration (BPA analysis without enzymatic deconjugation) and total BPA concentration (BPA analysis following deconjugation with -glucuronidase/sulfatase) [30,31]. Whereas glucuronides are major conjugated metabolites of BPA and TCS in urine, other minor conjugates include sulfates and glucuronide/sulfate derivatives [32]. To our knowledge, only Völkel et al. [33] developed an isotope dilution (ID)-LC–MS/MS method to specifically measure free BPA and BPAG in serum and urine samples, but their method lacked sensitivity due to interfering peaks. Using this method, they could not detect free BPA in any urine samples collected from 19 human subjects without known exposure to BPA. Most of these samples did contain detectable amounts BPAG but concentrations were always below the limit of quantification of 65 nmol L−1 (26.3 g L−1 ) [33]. Recently, Liao and Kannan [34] published a more sensitive LC–MS/MS method to analyze serum and urine samples for free BPA, BPA chlorides and BPA conjugates (BPAG, BPADS). Analysis of urine samples from 31 volunteers in Albany, New York, revealed that 96.8% and 87.1% of samples contained detectable amounts of free BPA and BPAG, respectively; corresponding geometric mean (GM) concentrations were 0.71 and 2.2 g L−1 . BPADS was also found above the LOQ in 35% of the samples (GM = 0.11 g L−1 ). A significant drawback of this method is the reliance on only one internal standard (13 C12 -BPA) for the quantification of all analytes. We developed methods for the determination of free BPA and TCS as well as their metabolites BPAS, BPADS, BPAG, TCSS, TCSG in urine samples, using ID-LC–MS/MS methods that comprised isotope-labeled standards for all analytes. The samples were also
analyzed for BPA and TCS (free and total) using a method combining enzymatic hydrolysis, derivatization with pentafluorobenzyl bromide, liquid–liquid extraction and GC–MS/MS, which was previously used in our laboratory. We also assessed possible losses of analytes during collection and storage of urine samples. 2. Experimental 2.1. Chemicals and reagents We purchased dansyl chloride, formic acid (>95%), ammonium formate (LC–MS Ultra), ß-glucuronidase from Helix pomatia and 2,3,4,5,6-pentafluorobenzyl bromide from Sigma-Aldrich (St. Louis, MO), acetone (pesticide grade), ammonium hydroxide (ACS – Pur reactive), hexane (Optima), sodium carbonate (ACS grade), anhydrous potassium carbonate (powder), acetic acid (glacial) and sodium acetate (ACS grade) from Fisher Scientific (Fairlawn, NJ), acetonitrile, methanol and dichloromethane (Omnisolv grade) from EMD (Omaha, NB), sodium bicarbonate (ACS grade) from JT Baker (Philipsburg, NJ) and synthetic urine (Surine Negative Urine Control) from Cerilliant (Round Rock, TX). Milli-Q water was purified by the Advantage A10 ultrapure water system (Merck Millipore, Billerica, MA). Chemical structures of native and isotope-labeled compounds analyzed in this study are shown in Fig. 1. BPA (>99%) and TCS (97%) were purchased from Sigma-Aldrich (St. Louis, MO) and used for calibration. BPA (98%) 100 g mL−1 and TCS (98%) 100 g mL−1 were bought from Cambridge Isotopes Laboratories (Andover, MA) and used for quality control. 13 C12 -BPA (99%) 100 g mL−1 and 13 C -TCS (99%) 100 g mL−1 were purchased from Cambridge Iso12 topes Laboratories. BPAG (98%) and BPAG-d6 (98%) were acquired from Toronto Research Chemicals (Toronto, ON, Canada). BPAS (>98%), BPADS (>98%), TCSS (>98%), TCSG (>98%), BPAS-d6 (>98%), BPADS-d6 (>98%), TCSS-d3 (97%) and TCSG-d3 (96%) were synthesized by the Organic Synthesis Service (Centre de recherche du CHU de Québec, Québec, QC, Canada). We also prepared dansylated derivatives of BPA, TCS, 13 C12 -BPA and 13 C12 -TCS for recovery assessment. All standard solutions were stored at −80 ◦ C until use. 2.2. Urine samples In March 2013, forty-six healthy volunteers each provided a spot urine sample collected in a 500-mL polypropylene bottle. A 10-mL aliquot was transferred into a 13 mL polypropylene tube and samples were stored frozen at −20 ◦ C until time of analysis. Donors (26 males; 20 females) were employees of the laboratory
Please cite this article in press as: G. Provencher, et al., J. Chromatogr. A (2014), http://dx.doi.org/10.1016/j.chroma.2014.04.072
G Model CHROMA-355364; No. of Pages 8
ARTICLE IN PRESS G. Provencher et al. / J. Chromatogr. A xxx (2014) xxx–xxx
and members of their family of ages ranging between 4 and 80 years (mean age = 34.5 years). 2.3. Analytical methods Two separate LC–MS/MS methods were developed: the first method is for free BPA and TCS quantification and the second for conjugate analysis. Total BPA and TCS concentrations in urine samples are compared to those obtained with a GC–MS/MS method, which is also described in this section and has been in use in our laboratory for several years. The procedures used for the validation of these methods by our laboratory accredited ISO/CEI 17025 are presented in Section A.1 (Appendix A). 2.3.1. Free BPA and TCS analysis by LC–MS/MS The stock solutions of BPA and TCS standards were prepared at 1000 mg L−1 in acetonitrile. The working standard solutions containing BPA and TCS were generated by serial dilution with acetonitrile. Stable isotope-labeled internal standards were purchased as solutions in acetonitrile and diluted in this solvent for the preparation of working solutions. All stock and working solutions were stored at −80 ◦ C until use. Spiking standard solutions were prepared in acetonitrile: H2 O (50:50, v:v) to cover concentration ranges from 0.01 to 2 g L−1 for BPA and from 0.025 to 10 g L−1 for TCS. Quality control (QC) materials were prepared from human urine obtained from volunteers in our laboratory. The urine, previously tested for its BPA and TCS content, was spiked at low and high concentrations by the addition of 0.18 g BPA L−1 and 0.9 g TCS L−1 , and 1.5 g BPA L−1 and 7.5 g TCS L−1 , respectively. The source of native compounds was different from that used for the preparation
3
of calibration curve standards. Aliquots of 2.25 mL were dispensed into glass vials and stored at −20 ◦ C until use. Urine and QC samples (1.0 mL) were transferred into 15-mL screw cap glass tubes. Calibration curve standards were spiked in water (1.0 mL) at each analytical day with 50 L of spiking standard solutions. Fifty L of internal standard solution (13 C12 -BPA- and 13 C -TCS) and 500 L of 0.5 M carbonate buffer were then added 12 and the samples vortex mixed for 1 min. A 2-mg dansyl chloride mL−1 solution in acetone (1.5 mL) was added and the samples vortex mixed again for 1 min. The samples were heated at 60 ◦ C for 20 min for compounds derivatization. After cooling, hexane (5 mL) was added to the samples. The tubes were vortex mixed for 3 min and centrifuged for 5 min at 2000 rpm. The upper organic phase was transferred to 16 mm × 100 mm disposable glass tubes and evaporated to dryness using a Zymark Turbo Vap (Zymark Corp., Hopkinton, MA) set at 40 ◦ C. The extracts were reconstituted in a solution of acetonitrile: H2 O (50:50, v:v) (2000 L). Analysis was performed on a Xevo TQ-S coupled with a UPLC system (Waters, Milford, MA). Ten microliters of extracted samples (standards, QC and urine samples) were injected onto an Acquity UPLC HSS T3, 1.8 m, 50 mm × 2.1 mm analytical column (Waters) maintained at 30 ◦ C. Elution was performed at a flow rate of 0.4 mL min−1 with a gradient starting from a 30:70 mixture of mobile phase A (0.1% aqueous formic acid solution) and mobile phase B (acetonitrile) at 0 min; increased linearly to 5:95 (A:B) from 0 to 4 min, held for 0.5 min; and finally reversed to 30:70 (A:B), held for 1 min before the next injection. The LC–MS/MS was operated in electrospray positive (ESI+) and multiple reaction monitoring (MRM) mode. The collision gas was argon at a flow rate of 0.15 mL min−1 and the cone energy was set to 50 eV. The retention times and mass transitions are presented in Table 1.
Table 1 Mass transitions monitored for bisphenol A (BPA), triclosan (TCS), their conjugates and isotope-labeled standards. Analyte
MW
ESI+ mode (dansylated derivatives) BPA 228.29 13
C12 -BPA
Retention time (min)
Mass transition(m/z)
Collision energy (eV)
4.08
695.5 > 171.2 695.5 > 156.2a 707.5 > 171.2 707.5 > 156.2a 522.3 > 171.2 522.3 > 156.2a 534.3 > 171.2 534.3 > 156.2a
48 74 48 74 30 50 30 50
307.3 > 212.3 307.3 > 227.2a 313.3 > 215.3 313.3 > 233.3a 387.2 > 227.3 387.2 > 212.3a 387.2 > 307.3a 393.2 > 233.4 403.4 > 212.3 403.4 > 227.3a 409.5 > 215.3 409.5 > 233.3a 366.8 > 286.8 368.6 > 288.8a 372.0 > 291.9 373.8 > 293.8a 462.8 > 289.2 465.1 > 289.2a 469.8 > 293.8 467.9 > 291.9a
30 24 30 24 36 40 20 36 40 34 40 34 15 15 15 15 14 14 14 14
240.20
4.07
TCS
289.54
3.44
13
301.45
3.44
C12 -TCS
ESI− mode BPAS
308.35
4.30
BPAS-d6
314.39
4.28
BPADS
388.41
3.35
BPADS-d6 BPAG
394.45 404.41
3.33 3.55
BPAG-d6
410.45
3.52
TCSS
369.60
6.02
TCSS-d3
372.62
6.02
TCSG
465.67
5.85
TCSG-d3
468.69
5.85
Ion ratiob
2.1 2.1 2.2 2.3
0.15 0.18 5.3 0.45 N/A 0.29 0.24 1.0 3.0 1.6 0.30
Abbreviations: BPA, bisphenol A; BPADS, bisphenol A disulfate; BPAG, bisphenol A glucuronide; BPAS, bisphenol A monosulfate; TCS, triclosan; TCSG, triclosan glucuronide; TCSS, triclosan sulfate. a Qualifier mass transition. b Quantifier/qualifier mass transition intensity ratio.
Please cite this article in press as: G. Provencher, et al., J. Chromatogr. A (2014), http://dx.doi.org/10.1016/j.chroma.2014.04.072
G Model CHROMA-355364; No. of Pages 8 4
ARTICLE IN PRESS G. Provencher et al. / J. Chromatogr. A xxx (2014) xxx–xxx
2.3.2. BPA and TCS conjugate analysis by LC–MS/MS Stock solutions of all the standards were prepared at 100 g mL−1 in MeOH:H2 O (50:50, v:v). Working standard solutions were generated by serial dilution with MeOH:H2 O (50:50, v:v). Stable isotope-labeled internal standards stock and working solutions were generated in MeOH:H2 O (50:50, v:v). All stock and working solutions were stored at −80 ◦ C until use. Synthetic urine (Surine) was used as the blank matrix to prepare our calibration curves. It was spiked daily with working standard solutions prepared in MeOH:H2 O (50:50, v:v) in the proportion of 50 L for 1.5 mL of matrix to cover concentrations ranging from 0.1 to 20 g L−1 for sulfate metabolites and from 0.1 to 100 g L−1 for glucuronide metabolites. QC materials were prepared in human urine obtained from volunteers in our laboratory. The urine, previously tested for its BPA and TCS metabolites content, was spiked at three different concentration levels: low (0.2 g L−1 ), medium (2 g L−1 for sulfate metabolites; 3 g L−1 for glucuronide metabolites) and high (15 g L−1 for sulfate metabolites; 60 g L−1 for glucuronide metabolites). Aliquots of 5 mL were dispensed into polypropylene tubes and stored at −20 ◦ C until use. A 1.5-mL aliquot of the preparation was used for sample analysis. Three replicates of each levels of quality control were analyzed in each analytical batch. Urine, QC and calibration standard samples (1.5 mL) were transferred into 16 mm × 100 mm disposable glass tubes. We then added 500 L of a 200 mM ammonium formate buffer solution (pH 4) and 20 L of the internal standard solution (BPAS-d6 , BPADS-d6 , BPAGd6 , TCSS-d3 and TCSG-d3 ). The sample was then vortex mixed and extracted by solid phase extraction (weak anion exchange mode) with a Strata X-AW extraction cartridge (100 mg/3 mL; Phenomenex, Torrance, CA). Using an extraction unit (Waters), we first conditioned the cartridge by gravity with 2 mL of methanol followed by 2 mL of Milli-Q type water. The flow was set to 1 mL min−1 during the loading, washing and elution steps. The sample was loaded on the conditioned cartridge which was then washed with 1 mL of a 10% MeOH in water solution, followed by 1 mL of 1% NH4 OH in a MeOH:H2 O (5:95, v:v) solution and then by 1 mL of the methanol solution. Analytes were eluted from the cartridge in a clean 16 mm × 100 mm glass tube using a 1% NH4 OH in methanol solution (1.5 mL). The solution containing the analytes was evaporated to dryness using a Zymark Turbo Vap set at 50 ◦ C. The extract was reconstituted in a solution of 25% methanol in water. Analysis was performed on the same instrument and column as described above. Ten microliters of extracted samples (standards, QC and urine samples) were injected onto the column maintained at 25 ◦ C. Elution was performed at a flow rate of 0.4 mL min−1 with a gradient starting from a 95:5 mixture of mobile phase A (2% NH4 OH in H2 O, pH 11) and mobile phase B (0.1% NH4 OH in methanol) at 0 min, held 0.5 min; increased linearly to 70:30 (A:B) from 0.5 to 3.5 min; increased linearly to 60:40 (A:B) from 3.5 to 4.0 min, held for 0.5 min; increased linearly to 40:60 (A:B) from 4.5 to 5.5 min; increased linearly to 0:100 (A:B) from 5.5 to 5.7 min, held 1 min; and linearly reversed to 95:5 (A:B) from 6.7 to 6.9 min, held for 1.1 min before the next injection. The LC–MS/MS was operated in electrospray negative (ESI−) and MRM mode. The collision gas was argon at a flow rate of 0.15 mL min−1 and the cone energy was set at 25 eV. Retention times and mass transitions are listed in Table 1. 2.3.3. BPA and TCS analysis by GC–MS/MS (total and free) The stock solutions of BPA and TCS standards were prepared at 1000 mg L−1 in acetonitrile. The working standard solutions, containing BPA and TCS were generated by serial dilution with acetonitrile. Stable isotope-labeled internal standards were purchased as solutions in acetonitrile and diluted in this solvent for the preparation of working solutions at 1 mg L−1 of 13 C12 -BPA and 10 mg L−1
of 13 C12 -TCS. All stock and working solutions were stored at −80 ◦ C until use. Spiking standard solutions were prepared in urine samples obtained from laboratory volunteers displaying low total BPA and TCS levels (
ARTICLE IN PRESS
CHROMA-355364; No. of Pages 8
Journal of Chromatography A, xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Determination of bisphenol A, triclosan and their metabolites in human urine using isotope-dilution liquid chromatography–tandem mass spectrometry Gilles Provencher a , René Bérubé a , Pierre Dumas a , Jean-Franc¸ois Bienvenu a , Éric Gaudreau a , Patrick Bélanger a , Pierre Ayotte a,b,c,∗ a
Centre de toxicologie du Québec, Institut national de santé publique du Québec (INSPQ), 945 Wolfe, Québec, QC, Canada G1V 5B3 Axe Santé des populations et pratiques optimales en santé, Centre de recherche du CHU de Québec, 2875 boul. Laurier, Édifice Delta 2, Bureau 600, Québec, QC, Canada G1V 2M2 c Département de médecine sociale et préventive, Université Laval, Pavillon Ferdinand-Vandry, Québec, QC, Canada G1V 0A6 b
a r t i c l e
i n f o
Article history: Received 30 December 2013 Received in revised form 15 April 2014 Accepted 20 April 2014 Available online xxx Keywords: Bisphenol A Triclosan Glucuronide metabolites Sulfate metabolites Isotope dilution LC–MS/MS Human urine
a b s t r a c t Bisphenol A (BPA) and triclosan (TCS) are ubiquitous environmental phenols exhibiting endocrine disrupting activities that may be involved in various health disorders in humans. There is a need to measure separately free forms and conjugated metabolites because only the former are biologically active. We have developed sensitive methods using isotope-dilution liquid chromatography–tandem mass spectrometry for individual measurements of free BPA and TCS as well as their metabolites, BPA glucuronide (BPAG), BPA monosulfate (BPAS), BPA disulfate (BPADS), TCS glucuronide (TCSG) and TCS sulfate (TCSS) in urine. Comparative analyses of urine samples from 46 volunteers living in the Quebec City area using the new methods and a GC–MS/MS method previously used in our laboratory revealed very strong correlations for total BPA (Spearman’s rs = 0.862, p < 0.0001) and total TCS concentrations (rs = 0.942, p < 0.0001). Glucuronide metabolites were the most abundant BPA and TCS species in urine samples (>94% of total urinary concentrations). Unconjugated TCS concentrations represented a small proportion of total TCS species (median = 1.6%) but its concentration was likely underestimated due to losses by adsorption to the surface of polypropylene tubes used for sample storage. To our knowledge, we are the first to report levels of free, sulfated and glucuronidated TCS levels in human urine. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Bisphenol A (BPA) and triclosan (TCS) are two widely used phenolic compounds with numerous industrial and commercial applications. BPA is the monomer in the production of polycarbonate plastics and is also a component of epoxy resins that are used to produce food and beverage packaging materials and dental sealants among others [1]. Trace amounts of BPA have been shown to leach from polycarbonate containers [2–4], which led to the ban of polycarbonate baby bottles in Canada in 2010. BPA has been found in foodstuffs [5,6], and diet is most likely the major source of exposure to this compound [1,7,8]. TCS is a bactericidal agent added to soap and toothpaste [9,10]. Dermal absorption and ingestion of TCS in
∗ Corresponding author at: Centre de toxicologie du Québec, Institut national de santé publique du Québec (INSPQ), 945 Wolfe, Québec, QC, Canada G1V 5B3. Tel.: +1 418 650 5115x4654; fax: +1 418 654 2148. E-mail address: [email protected] (P. Ayotte).
these personal care products are thought to contribute the most to human exposure [9]. Both BPA and TCS are endocrine disrupting chemicals: BPA is an estrogen receptor agonist and an androgen receptor antagonist [11,12]. TCS interacts with constitutive androstane and pregnane-X receptors [13] and exhibits antagonistic activity in both estrogen receptor- and androgen receptor-responsive bioassays [14]. Human exposure to these compounds have been associated with various adverse health effects including reproductive and endocrine disorders, cardiovascular disease, obesity and metabolic syndrome, cancer, immune system dysfunction and neurobehavioral defects [1,15–17]. However, whether or not these associations indicate an important public health risk is highly controversial. An argument raised by those supporting a trivial health risk from exposure to these phenolic compounds is their rapid biotransformation to biologically-inactive metabolites [18]. Indeed, following their absorption by the gastro-intestinal tract, free BPA and TCS are rapidly and efficiently conjugated by hepatic uridine diphosphate-glucuronyltransferases to their glucuronidated forms
http://dx.doi.org/10.1016/j.chroma.2014.04.072 0021-9673/© 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: G. Provencher, et al., J. Chromatogr. A (2014), http://dx.doi.org/10.1016/j.chroma.2014.04.072
G Model CHROMA-355364; No. of Pages 8 2
ARTICLE IN PRESS G. Provencher et al. / J. Chromatogr. A xxx (2014) xxx–xxx
Fig. 1. Chemical structures of native and isotope-labeled compounds analyzed in this study. Internal standards for free BPA (13 C12 -BPA) and free TCS (13 C12 -TCS) bear 13 C atoms at all positions of aromatic rings.
and eliminated in urine [19,20]. However, the recent demonstration in the dog model of an efficient and rapid absorption of BPA through the oral mucosa by the sublingual route (avoiding the first-pass hepatic metabolism) could indicate a greater systemic exposure to free BPA by the oral route than previously thought [21]. Also, BPAG can be deconjugated by -glucuronidase found in placenta, liver, kidney, and intestine [18]. The development of an analytical method allowing the determination of unconjugated (“free”) and conjugated forms of BPA and TCS is central to an improved assessment of the health risk associated with exposure to these ubiquitous compounds. Both liquid chromatography–tandem mass spectrometry (LC–MS/MS) and gas chromatography–mass spectrometry (GC–MS) techniques have been used for the measurement of BPA and its metabolites in biological specimens [22–24]. Because standards of BPA conjugates were not commercially available, most methods measured only total BPA and TCS concentrations following enzymatic hydrolysis of conjugates [25–29]. Some researchers measured both free BPA concentration (BPA analysis without enzymatic deconjugation) and total BPA concentration (BPA analysis following deconjugation with -glucuronidase/sulfatase) [30,31]. Whereas glucuronides are major conjugated metabolites of BPA and TCS in urine, other minor conjugates include sulfates and glucuronide/sulfate derivatives [32]. To our knowledge, only Völkel et al. [33] developed an isotope dilution (ID)-LC–MS/MS method to specifically measure free BPA and BPAG in serum and urine samples, but their method lacked sensitivity due to interfering peaks. Using this method, they could not detect free BPA in any urine samples collected from 19 human subjects without known exposure to BPA. Most of these samples did contain detectable amounts BPAG but concentrations were always below the limit of quantification of 65 nmol L−1 (26.3 g L−1 ) [33]. Recently, Liao and Kannan [34] published a more sensitive LC–MS/MS method to analyze serum and urine samples for free BPA, BPA chlorides and BPA conjugates (BPAG, BPADS). Analysis of urine samples from 31 volunteers in Albany, New York, revealed that 96.8% and 87.1% of samples contained detectable amounts of free BPA and BPAG, respectively; corresponding geometric mean (GM) concentrations were 0.71 and 2.2 g L−1 . BPADS was also found above the LOQ in 35% of the samples (GM = 0.11 g L−1 ). A significant drawback of this method is the reliance on only one internal standard (13 C12 -BPA) for the quantification of all analytes. We developed methods for the determination of free BPA and TCS as well as their metabolites BPAS, BPADS, BPAG, TCSS, TCSG in urine samples, using ID-LC–MS/MS methods that comprised isotope-labeled standards for all analytes. The samples were also
analyzed for BPA and TCS (free and total) using a method combining enzymatic hydrolysis, derivatization with pentafluorobenzyl bromide, liquid–liquid extraction and GC–MS/MS, which was previously used in our laboratory. We also assessed possible losses of analytes during collection and storage of urine samples. 2. Experimental 2.1. Chemicals and reagents We purchased dansyl chloride, formic acid (>95%), ammonium formate (LC–MS Ultra), ß-glucuronidase from Helix pomatia and 2,3,4,5,6-pentafluorobenzyl bromide from Sigma-Aldrich (St. Louis, MO), acetone (pesticide grade), ammonium hydroxide (ACS – Pur reactive), hexane (Optima), sodium carbonate (ACS grade), anhydrous potassium carbonate (powder), acetic acid (glacial) and sodium acetate (ACS grade) from Fisher Scientific (Fairlawn, NJ), acetonitrile, methanol and dichloromethane (Omnisolv grade) from EMD (Omaha, NB), sodium bicarbonate (ACS grade) from JT Baker (Philipsburg, NJ) and synthetic urine (Surine Negative Urine Control) from Cerilliant (Round Rock, TX). Milli-Q water was purified by the Advantage A10 ultrapure water system (Merck Millipore, Billerica, MA). Chemical structures of native and isotope-labeled compounds analyzed in this study are shown in Fig. 1. BPA (>99%) and TCS (97%) were purchased from Sigma-Aldrich (St. Louis, MO) and used for calibration. BPA (98%) 100 g mL−1 and TCS (98%) 100 g mL−1 were bought from Cambridge Isotopes Laboratories (Andover, MA) and used for quality control. 13 C12 -BPA (99%) 100 g mL−1 and 13 C -TCS (99%) 100 g mL−1 were purchased from Cambridge Iso12 topes Laboratories. BPAG (98%) and BPAG-d6 (98%) were acquired from Toronto Research Chemicals (Toronto, ON, Canada). BPAS (>98%), BPADS (>98%), TCSS (>98%), TCSG (>98%), BPAS-d6 (>98%), BPADS-d6 (>98%), TCSS-d3 (97%) and TCSG-d3 (96%) were synthesized by the Organic Synthesis Service (Centre de recherche du CHU de Québec, Québec, QC, Canada). We also prepared dansylated derivatives of BPA, TCS, 13 C12 -BPA and 13 C12 -TCS for recovery assessment. All standard solutions were stored at −80 ◦ C until use. 2.2. Urine samples In March 2013, forty-six healthy volunteers each provided a spot urine sample collected in a 500-mL polypropylene bottle. A 10-mL aliquot was transferred into a 13 mL polypropylene tube and samples were stored frozen at −20 ◦ C until time of analysis. Donors (26 males; 20 females) were employees of the laboratory
Please cite this article in press as: G. Provencher, et al., J. Chromatogr. A (2014), http://dx.doi.org/10.1016/j.chroma.2014.04.072
G Model CHROMA-355364; No. of Pages 8
ARTICLE IN PRESS G. Provencher et al. / J. Chromatogr. A xxx (2014) xxx–xxx
and members of their family of ages ranging between 4 and 80 years (mean age = 34.5 years). 2.3. Analytical methods Two separate LC–MS/MS methods were developed: the first method is for free BPA and TCS quantification and the second for conjugate analysis. Total BPA and TCS concentrations in urine samples are compared to those obtained with a GC–MS/MS method, which is also described in this section and has been in use in our laboratory for several years. The procedures used for the validation of these methods by our laboratory accredited ISO/CEI 17025 are presented in Section A.1 (Appendix A). 2.3.1. Free BPA and TCS analysis by LC–MS/MS The stock solutions of BPA and TCS standards were prepared at 1000 mg L−1 in acetonitrile. The working standard solutions containing BPA and TCS were generated by serial dilution with acetonitrile. Stable isotope-labeled internal standards were purchased as solutions in acetonitrile and diluted in this solvent for the preparation of working solutions. All stock and working solutions were stored at −80 ◦ C until use. Spiking standard solutions were prepared in acetonitrile: H2 O (50:50, v:v) to cover concentration ranges from 0.01 to 2 g L−1 for BPA and from 0.025 to 10 g L−1 for TCS. Quality control (QC) materials were prepared from human urine obtained from volunteers in our laboratory. The urine, previously tested for its BPA and TCS content, was spiked at low and high concentrations by the addition of 0.18 g BPA L−1 and 0.9 g TCS L−1 , and 1.5 g BPA L−1 and 7.5 g TCS L−1 , respectively. The source of native compounds was different from that used for the preparation
3
of calibration curve standards. Aliquots of 2.25 mL were dispensed into glass vials and stored at −20 ◦ C until use. Urine and QC samples (1.0 mL) were transferred into 15-mL screw cap glass tubes. Calibration curve standards were spiked in water (1.0 mL) at each analytical day with 50 L of spiking standard solutions. Fifty L of internal standard solution (13 C12 -BPA- and 13 C -TCS) and 500 L of 0.5 M carbonate buffer were then added 12 and the samples vortex mixed for 1 min. A 2-mg dansyl chloride mL−1 solution in acetone (1.5 mL) was added and the samples vortex mixed again for 1 min. The samples were heated at 60 ◦ C for 20 min for compounds derivatization. After cooling, hexane (5 mL) was added to the samples. The tubes were vortex mixed for 3 min and centrifuged for 5 min at 2000 rpm. The upper organic phase was transferred to 16 mm × 100 mm disposable glass tubes and evaporated to dryness using a Zymark Turbo Vap (Zymark Corp., Hopkinton, MA) set at 40 ◦ C. The extracts were reconstituted in a solution of acetonitrile: H2 O (50:50, v:v) (2000 L). Analysis was performed on a Xevo TQ-S coupled with a UPLC system (Waters, Milford, MA). Ten microliters of extracted samples (standards, QC and urine samples) were injected onto an Acquity UPLC HSS T3, 1.8 m, 50 mm × 2.1 mm analytical column (Waters) maintained at 30 ◦ C. Elution was performed at a flow rate of 0.4 mL min−1 with a gradient starting from a 30:70 mixture of mobile phase A (0.1% aqueous formic acid solution) and mobile phase B (acetonitrile) at 0 min; increased linearly to 5:95 (A:B) from 0 to 4 min, held for 0.5 min; and finally reversed to 30:70 (A:B), held for 1 min before the next injection. The LC–MS/MS was operated in electrospray positive (ESI+) and multiple reaction monitoring (MRM) mode. The collision gas was argon at a flow rate of 0.15 mL min−1 and the cone energy was set to 50 eV. The retention times and mass transitions are presented in Table 1.
Table 1 Mass transitions monitored for bisphenol A (BPA), triclosan (TCS), their conjugates and isotope-labeled standards. Analyte
MW
ESI+ mode (dansylated derivatives) BPA 228.29 13
C12 -BPA
Retention time (min)
Mass transition(m/z)
Collision energy (eV)
4.08
695.5 > 171.2 695.5 > 156.2a 707.5 > 171.2 707.5 > 156.2a 522.3 > 171.2 522.3 > 156.2a 534.3 > 171.2 534.3 > 156.2a
48 74 48 74 30 50 30 50
307.3 > 212.3 307.3 > 227.2a 313.3 > 215.3 313.3 > 233.3a 387.2 > 227.3 387.2 > 212.3a 387.2 > 307.3a 393.2 > 233.4 403.4 > 212.3 403.4 > 227.3a 409.5 > 215.3 409.5 > 233.3a 366.8 > 286.8 368.6 > 288.8a 372.0 > 291.9 373.8 > 293.8a 462.8 > 289.2 465.1 > 289.2a 469.8 > 293.8 467.9 > 291.9a
30 24 30 24 36 40 20 36 40 34 40 34 15 15 15 15 14 14 14 14
240.20
4.07
TCS
289.54
3.44
13
301.45
3.44
C12 -TCS
ESI− mode BPAS
308.35
4.30
BPAS-d6
314.39
4.28
BPADS
388.41
3.35
BPADS-d6 BPAG
394.45 404.41
3.33 3.55
BPAG-d6
410.45
3.52
TCSS
369.60
6.02
TCSS-d3
372.62
6.02
TCSG
465.67
5.85
TCSG-d3
468.69
5.85
Ion ratiob
2.1 2.1 2.2 2.3
0.15 0.18 5.3 0.45 N/A 0.29 0.24 1.0 3.0 1.6 0.30
Abbreviations: BPA, bisphenol A; BPADS, bisphenol A disulfate; BPAG, bisphenol A glucuronide; BPAS, bisphenol A monosulfate; TCS, triclosan; TCSG, triclosan glucuronide; TCSS, triclosan sulfate. a Qualifier mass transition. b Quantifier/qualifier mass transition intensity ratio.
Please cite this article in press as: G. Provencher, et al., J. Chromatogr. A (2014), http://dx.doi.org/10.1016/j.chroma.2014.04.072
G Model CHROMA-355364; No. of Pages 8 4
ARTICLE IN PRESS G. Provencher et al. / J. Chromatogr. A xxx (2014) xxx–xxx
2.3.2. BPA and TCS conjugate analysis by LC–MS/MS Stock solutions of all the standards were prepared at 100 g mL−1 in MeOH:H2 O (50:50, v:v). Working standard solutions were generated by serial dilution with MeOH:H2 O (50:50, v:v). Stable isotope-labeled internal standards stock and working solutions were generated in MeOH:H2 O (50:50, v:v). All stock and working solutions were stored at −80 ◦ C until use. Synthetic urine (Surine) was used as the blank matrix to prepare our calibration curves. It was spiked daily with working standard solutions prepared in MeOH:H2 O (50:50, v:v) in the proportion of 50 L for 1.5 mL of matrix to cover concentrations ranging from 0.1 to 20 g L−1 for sulfate metabolites and from 0.1 to 100 g L−1 for glucuronide metabolites. QC materials were prepared in human urine obtained from volunteers in our laboratory. The urine, previously tested for its BPA and TCS metabolites content, was spiked at three different concentration levels: low (0.2 g L−1 ), medium (2 g L−1 for sulfate metabolites; 3 g L−1 for glucuronide metabolites) and high (15 g L−1 for sulfate metabolites; 60 g L−1 for glucuronide metabolites). Aliquots of 5 mL were dispensed into polypropylene tubes and stored at −20 ◦ C until use. A 1.5-mL aliquot of the preparation was used for sample analysis. Three replicates of each levels of quality control were analyzed in each analytical batch. Urine, QC and calibration standard samples (1.5 mL) were transferred into 16 mm × 100 mm disposable glass tubes. We then added 500 L of a 200 mM ammonium formate buffer solution (pH 4) and 20 L of the internal standard solution (BPAS-d6 , BPADS-d6 , BPAGd6 , TCSS-d3 and TCSG-d3 ). The sample was then vortex mixed and extracted by solid phase extraction (weak anion exchange mode) with a Strata X-AW extraction cartridge (100 mg/3 mL; Phenomenex, Torrance, CA). Using an extraction unit (Waters), we first conditioned the cartridge by gravity with 2 mL of methanol followed by 2 mL of Milli-Q type water. The flow was set to 1 mL min−1 during the loading, washing and elution steps. The sample was loaded on the conditioned cartridge which was then washed with 1 mL of a 10% MeOH in water solution, followed by 1 mL of 1% NH4 OH in a MeOH:H2 O (5:95, v:v) solution and then by 1 mL of the methanol solution. Analytes were eluted from the cartridge in a clean 16 mm × 100 mm glass tube using a 1% NH4 OH in methanol solution (1.5 mL). The solution containing the analytes was evaporated to dryness using a Zymark Turbo Vap set at 50 ◦ C. The extract was reconstituted in a solution of 25% methanol in water. Analysis was performed on the same instrument and column as described above. Ten microliters of extracted samples (standards, QC and urine samples) were injected onto the column maintained at 25 ◦ C. Elution was performed at a flow rate of 0.4 mL min−1 with a gradient starting from a 95:5 mixture of mobile phase A (2% NH4 OH in H2 O, pH 11) and mobile phase B (0.1% NH4 OH in methanol) at 0 min, held 0.5 min; increased linearly to 70:30 (A:B) from 0.5 to 3.5 min; increased linearly to 60:40 (A:B) from 3.5 to 4.0 min, held for 0.5 min; increased linearly to 40:60 (A:B) from 4.5 to 5.5 min; increased linearly to 0:100 (A:B) from 5.5 to 5.7 min, held 1 min; and linearly reversed to 95:5 (A:B) from 6.7 to 6.9 min, held for 1.1 min before the next injection. The LC–MS/MS was operated in electrospray negative (ESI−) and MRM mode. The collision gas was argon at a flow rate of 0.15 mL min−1 and the cone energy was set at 25 eV. Retention times and mass transitions are listed in Table 1. 2.3.3. BPA and TCS analysis by GC–MS/MS (total and free) The stock solutions of BPA and TCS standards were prepared at 1000 mg L−1 in acetonitrile. The working standard solutions, containing BPA and TCS were generated by serial dilution with acetonitrile. Stable isotope-labeled internal standards were purchased as solutions in acetonitrile and diluted in this solvent for the preparation of working solutions at 1 mg L−1 of 13 C12 -BPA and 10 mg L−1
of 13 C12 -TCS. All stock and working solutions were stored at −80 ◦ C until use. Spiking standard solutions were prepared in urine samples obtained from laboratory volunteers displaying low total BPA and TCS levels (
