Quantification of urinary conjugates of bisphenol A, 2,5-dichlorophenol, and 2-hydroxy-4-methoxybenzophenone in humans by online solid phase extraction–high performance liquid chromatography–tandem mass spectrometry
Urinary concentrations of phenols or their metabolites have been used as biomarkers to assess the prevalence of exposure to these compounds in the general population. Total urinary concentrations, which include both free and conjugated (glucuronide and sulfated) forms of the compounds, are usually reported. From a toxicologic standpoint, the relative concentrations of the free species compared with their conjugated analogs can be important because conjugation may reduce the potential biologic activity of the phenols. In this study, we determined the percentage of glucuronide and sulfate conjugates of three phenolic compounds, bisphenol A (BPA), 2,5-dichlorophenol (2,5-DCP), and 2-hydroxy-4-methoxybenzophenone (benzophenone-3, BP-3) in 30 urine samples collected between 2000 and 2004 from a demographically diverse group of anonymous adult volunteers. We used a sensitive on-line solid phase extraction–isotope dilution–high performance liquid chromatography–tandem mass spectrometry method. These three phenols were detected frequently in the urine samples tested. Only small percentages of the compounds (9.5% for BPA, and 3% for 2,5-DCP and BP-3) were excreted in their free form. The percentage of the sulfate conjugate was about twice that of the free compound. The glucuronide conjugate was the major metabolite, representing 69.5% (BPA), 89% (2,5-DCP), and 84.6% (BP-3) of the total amount excreted in urine. These results are in agreement with those reported before which suggested that BPA-glucuronide was an important BPA urinary metabolite in humans. To our knowledge, this is the first study describing the distribution of urinary conjugates of BP-3 and 2,5-DCP in humans.
1,4-Dichlorobenzene (1,4-DCB) is used as a moth repellent and a deodorant in household and industry; it is also an intermediate in the production of pesticides, dyestuff, and drugs . The genotoxic potential of 1,4-DCB has been evaluated extensively in vitro and in vivo [13–15]. 1,4-DCB caused tumors in kidneys of male rats and in livers of male and female mice. Because 1,4-DCB is widely used in the United States (the estimated production was 65 million kg in 1990) , environmental exposure to 1,4-DCB in humans could be substantial. In mammals, 1,4-DCB is mainly hydroxylated to 2,5-dichlorophenol (2,5-DCP, Fig. 1), and 2,5-DCP can then be conjugated with β-D-glucuronide and sulfate and excreted in the urine [14, 16]. 2,5-DCP has been used as a biomarker of expose to 1,4-DCB [17, 18]. 2,5-DCP has also been used as an intermediate in the production of dyes, pharmaceutical, and agricultural products such as insecticides (clothes moths, fruit borers, and ants); fumigants (pig stalls); germicides (garbage, restrooms); lubricants; and to control mildew and mold (leather and fabrics). In the United States, the total concentration of 2,5-DCP was measured in 982 urine samples collected during 1988–1994 from adult participants in the Third National Health and Nutrition Examination Survey (NHANES III) . The frequency of detection of 2,5-DCP was 98%, and the median and 95th percentile concentrations were 30 μg/L (24 μg/g creatinine), and 790 μg/L (670 μg/g creatinine), respectively. More recently, 2,5-DCP was detected in 1,989 urine samples collected in the United States during 1999–2000 from participants between 6 and 59 years of age in NHANES 1999–2000 . The geometric mean and median concentrations were 6.01 μg/L (5.38 μg per gram of creatinine [μg/g creatinine]) and 6.50 μg/L (5.60 μg/g creatinine), respectively; the 95th percentile concentration was 440 μg/L (299 μg/g creatinine). 2,5-DCP median concentrations from NHANES III were five times higher than concentrations from NHANES 1999–2000. These findings may reflect reduced exposures to 2,5-DCP or its precursors since the NHANES III sampling or differences among sample populations.
2-Hydroxy-4-methoxybenzophenone (benzophenone-3, BP-3, Fig. 1) is a common ingredient in sunscreens and other personal care products because of its ability to absorb and dissipate ultraviolet light. BP-3 can exert a uterotrophic effect in vivo, stimulate cell proliferation of MCF-7 breast cancer cells, and increase the secretion of tumor marker pS2 in vitro . In rats and humans, the metabolism of BP-3 produces 2,4-dihydroxylbenzophenone (DHB), 2,2′-dihydroxy-4-methoxybenzophenone (DHMB), and 2,3,4-trihydroxybenzophenone (THB) [20–22]. BP-3 and its metabolites can be conjugated with β-D-glucuronide and sulfate, and excreted in urine. Therefore, BP-3 or any of its metabolites can potentially be used as biomarkers of exposure to BP-3. To date, little information is available on the levels of BP-3 or its metabolites in humans.
Conjugation with β-D-glucuronide and sulfate may reduce the bioactivity of chemicals while facilitating their urinary excretion. Therefore, determining the relative concentrations of the conjugated and free species may be important for risk assessment . However, identifying and measuring the conjugated species of xenobiotics may be challenging because conjugated standards are not always readily available, and sensitive and accurate analytical methods are required to measure the concentrations of these species at trace levels. Gas chromatography (GC) coupled with mass spectrometry (MS) has been used extensively for measuring phenols in urine [24–28]. We have developed a novel highly sensitive on-line solid phase extraction–isotope dilution–high performance liquid chromatography–tandem mass spectrometry (on-line SPE–HPLC–MS/MS) method to determine BPA, 2,5-DCP, and BP-3 in urine. Compared with previous GC–MS methods, on-line SPE–HPLC–MS/MS provides comparable sensitivity, precision, and accuracy, but eliminates the usually complicated and time-consuming sample preparation steps, such as derivatization, preconcentration, and reconstitution. In the present work, we used a selective enzymatic hydrolysis treatment followed by on-line SPE–HPLC–MS/MS to determine the free, glucuronide, and sulfate forms of BPA, 2,5-DCP, and BP-3 in 30 urine samples collected between 2000 and 2004 from a convenience group of demographically diverse anonymous adult volunteers. To our knowledge, this is the first study describing the distribution of urinary conjugates of BP-3 and 2,5-DCP in humans.
Methanol (MeOH) and water, purchased from Caledon Laboratory Inc. (Ontario, Canada) were analytic or HPLC grade. Ammonium acetate (NH4Ac), BPA, 2,5-DCP, 4-methylumbelliferyl glucuronide (4-UMB-glu), 4-methylumbelliferyl sulfate (potassium salt) (4-UMB-sul), 4-methylumbelliferone (4-UMB), and β-glucuronidase (Helix pomatia, H1) were purchased from Sigma Aldrich Laboratories, Inc. (St. Louis, MO, USA). BP-3 was provided by EMD Chemicals Inc. (Hawthorne, NY, USA). Arylsulfatase (Helix pomatia) and β-glucuronidase (Escherichia coli-K12) were purchased from Roche Applied Science (Penzberg, Germany). 13C12-BPA, 13C6-2,5-DCP, and 13C4-4-UMB were obtained from Cambridge Isotope Laboratories Inc. (Andover, MA, USA).
Preparation of standard solutions and quality control materials
Ten working standard spiking solutions, containing BPA 2,5-DCP, and BP-3, were generated by serial dilution with MeOH of the initial stock solutions to cover concentration ranges of 0.1–1,000 μg/L (2,5-DCP), 0.1–100 μg/L (BPA), and 0.1–200 μg/L (BP-3). The isotope-labeled standard spiking solution was also prepared in MeOH. Quality control (QC) materials were prepared from a urine pool obtained from multiple anonymous adult donors. The urine pool was divided into two for QC low (QCL) and QC high (QCH) concentration pools. The QCL and the QCH pools were enriched with different levels of the native target compounds. All QC materials were stored at −20 °C. The QC pools were characterized to define the mean and the 95% and 99% control limits of BPA, 2,5-DCP, and BP-3 concentrations by a minimum of 40 repeated measurements in a five week period.
A stock solution of 4-UMB conjugates, prepared by dissolving 4-UMB-glu and 4-UMB-sul in MeOH, was diluted with MeOH to generate a standard solution containing 20 μg/mL of both 4-UMB-glu and 4-UMB-sul. Ten microliters of this solution was added to all samples. The 4-UMB/13C4-4-UMB peak area ratio was monitored to check the extent of the deconjugation reaction.
The β-glucuronidase (E. coli) solution was prepared by diluting 100 μL of enzyme to 5 mL with 1M NH4Ac buffer (pH 6.5). The β-glucuronidase (H. pomatia) solution was prepared by dissolving 0.01 grams of the enzyme with 5 mL of 1 M NH4Ac buffer (pH 5.0).
All unknown, blank, calibration standard, and QC samples were prepared in autosampler vials. Unknown and blank samples were prepared in three different ways: 1) processed without enzyme to determine the free species, 2) treated with β-glucuronidase (E. coli) to determine the free plus glucuronidated species, or 3) treated with β-glucuronidase (H. pomatia) to determine the free plus glucuronidated plus sulfated species. A 100 μL aliquot of urine was mixed with 50 μL of internal standard solution, 10 μL of 4-UMB-glu/4-UMB-sul solution, and 50 μL of the appropriate enzyme (or 1 M NH4Ac buffer [pH 5.0] for the analyses with no enzyme treatment). Calibration standards and blanks were prepared using 100 μL of the native spiking solution and 100 μL of water instead of urine, respectively. After gentle mixing, the samples were incubated at 37 °C overnight, then diluted with 790 μL of formic acid, and centrifuged.
The on-line SPE-HPLC-MS/MS system consisted of several Agilent 1,100 modules (Agilent Technologies, Wilmington, DE, USA), namely two binary pumps with degassers, an autosampler with a 900 μL injection loop, and one column compartment with a six-port switching valve, coupled with an API 4000 triple quadrupole mass spectrometer (Applied Biosystems, Foster City, CA, USA) equipped with an atmospheric pressure chemical ionization (APCI) interface. The mass spectrometer and Agilent modules were programmed and controlled using the Analyst 1.4 software (Applied Biosystems).
Negative ion APCI was used to generate gas phase ions. The APCI settings were curtain gas (N2) flow: 20 arbitrary units (au), collision gas flow: 9 au, nebulizer gas (air) flow: 50 au, nebulizing gas temperature: 500 °C, and corona needle voltage: −3 V. Ionization parameters and collision cell parameters were optimized separately for each analyte. Unit resolution was used for both Q1 and Q3 quadrupoles. We monitored two precursor/product ion transitions (m/z) for each analyte and one for the corresponding isotopically labeled analog as follows: BPA (227/133, 227/212), 13C12-BPA (239/139), 2,5-DCP (161/125, 163/125), 13C6-2,5-DCP (167/131), and BP-3 (227/183, 227/221). We used 13C12-BPA as the internal standard for BP-3. For 4-methylumbelliferone and 13C4-4-methylumbelliferone, we used the m/z transitions 175/133 and 179/135, respectively.
Analytical separation and quantification of 4-UMB conjugates during enzymatic hydrolysis
We followed the formation of 4-UMB from 4-UMB conjugates as a function of incubation time by HPLC-MS/MS. 4-UMB, 4-UMB-glu, and 4-UMB-sul were separated using a Prism RP (5 μm, 50×3 mm, Thermo Electron Corp., San Jose, CA, USA) HPLC column. The mobile phase A was 10 mM ammonium acetate (pH 6.5) and the mobile phase B was MeOH:acetonitrile (50:50). The gradient program was 15% MeOH/H2O for 5 min at a flow rate of 0.8 mL/min, increase to 50% MeOH/H2O at 5.1 min, keep 50% MeOH/H2O from 5.1 min to 8 min, decrease to 15% MeOH/H2O at 8.1 min, and keep for 2 min to allow for column re-equilibration.
To select the mass transitions for the quantification of the different 4-UMB species during the enzymatic hydrolysis, full-scan molecular ion and product ion spectra of the 4-UMB, 4-UMB-glu, and 4-UMB-sul were obtained from infusion of each compound at a flow rate of 10 μL/min. The full scan showed strong [M-H]− signals at m/z 175 (4-UMB), 351 (4-UMB-glu), and 225 (4-UMB-sul). The product ion spectra had strong transitions from m/z 175 to m/z 133 (4-UMB), m/z 351 to m/z 175 (4-UMB-glu, loss of the glucuronic acid), and m/z 225 to m/z 175 (4-UMB-sul, loss of [SO3]). We used these transitions for the quantification of 4-UMB, 4-UMB-glu, and 4-UMB-sul, respectively, at the retention times for each compound. Baseline separation of the free and conjugated 4-UMB species was important because the extracted ion chromatogram of 4-UMB showed two additional small peaks at the retention times of 4-UMB-glu and 4-UMB-sul. These signals correspond to 4-UMB resulting from the dissociation of 4-UMB-glu and 4-UMB-sul in the LC/MS interface when using the relatively harsh APCI ionization mode.
Results and discussion
Optimizing the selective enzymatic treatment of 4-UMB conjugates
We evaluated the activity of three commercial enzymes, β-glucuronidase (Escherichia coli), arylsulfatase (Helix pomatia), and β-glucuronidase (Helix pomatia), on 4-UMB-sul and 4-UMB-glu as the substrates. We treated 10 mL of 4-UMB-glu and 4-UMB-sul standard solution (2 μg/mL) separately with and without enzyme at 37 °C. The optimal amount and pH conditions for each enzyme were as follows: 50 μL of E. coli β-glucuronidase at pH 6.5, 200 μL of arylsulfatase at pH 6.5, and 0.5 mg of H. pomatia β-glucuronidase at pH 5.
Validating the on-line SPE-HPLC-MS/MS method
Spiked standard concentration recoveries, limits of detection (LODs), and precision of concentration measurements in spiked quality control (QC) samples for bisphenol A (BPA), 2,5-dichlorophenol (2,5-DCP), and benzophenone-3 (BP-3)
(Standard concentration in μg/L) Spiked recovery (%)
Determination of the urinary conjugates of BPA, 2,5-DCP, and BP-3
We calculated the percentage of glucuronide and sulfate conjugates of BPA, 2,5-DCP and BP-3 in 30 urine samples collected between 2000 and 2004 in Atlanta, GA (USA) from a convenience sample of a demographically diverse group of anonymous adult volunteers with no known occupational exposure to these phenols. The urine samples were collected at different times throughout the day and were not necessarily first-morning voids. To determine the concentrations of free species (Cfree), we performed the analyses without enzyme treatment. We used β-glucuronidase (E. coli) to determine the free plus glucuronidated species (Cfree+glu), and β-glucuronidase (H. pomatia) to determine the free plus glucuronidated plus sulfated species (Cfree+glu+sul). For each analyte, we calculated the concentration of glucuronide conjugates by subtracting Cfree from Cglu+free. Similarly, we calculated the concentration of sulfate metabolites by subtracting Cfree+glu+sul from Cfree+glu.
Urinary concentrations (in micrograms per liter of urine, μg/L) of the free, glucuronidated, and sulfated conjugates of bisphenol A (BPA), 2,5-dichlorophenol (2,5-DCP), and benzophenone-3 (BP-3)
Frequency of detection (%)
Percentage of total amount (%)
The present results for BPA are in agreement with those reported before which suggested that BPA-glucuronide was the main BPA urinary metabolite following the low dose oral administration of BPA to six persons . Our study also showed that the percentages of free 2,5-DCP and BP-3 were very similar and about three times lower than the percentage of free BPA. We speculate that BPA (the only dihydroxy compound of the three measured in this study) is relatively more hydrophilic than the others, thus explaining the higher percentage of BPA excreted as a free species compared with BP-3 and 2,5-DCP. Gender differences in the levels of urinary BPA conjugates were reported in a group of 30 healthy Korean adults, with men having significantly higher levels of the glucuronide than women, but women had higher levels of the sulfate . The percentage of the total BPA excreted in its free form was similar for both men and women (~30%) . Although no demographic information was available to evaluate potential gender differences, the percentage of BPA excreted in its free form in our study population is about one third of the free BPA excreted by the Korean adults. Reasons for these differences are, at present, unknown.
Human exposure to BP-3 has been investigated using BP-3 and its metabolites, including DHB and DHMB as biomarkers in urine . To our knowledge, this is the first study describing the distribution of glucuronide and sulfate conjugates of BP-3 in human urine. Similarly, no published studies are available on the distribution of 2,5-DCP urinary conjugates in humans. In male Wistar rats, the main urinary conjugate of 2,5-DCP was the sulfate (50%–60%), followed by the glucuronide (20%–30%) and the free form (5%–10%) . In agreement with the animal data, our results support the notion that 2,5-DCP is mainly excreted in the urine in a conjugated form. In rats the sulfate was the main conjugate, but in humans we found that the glucuronide is the main species. Although the metabolism of xenobiotics in animals and humans often results in similar metabolites, differences among species, especially in the distribution of the metabolites, are not uncommon. For instance, although ~80% of the urinary metabolites of di(2-ethylhexyl) phthalate (DEHP) in monkeys and humans are excreted in the form of glucuronide conjugates, urinary DEHP metabolites are excreted unconjugated in rats .
In summary, our results suggest that BPA, 2,5-DCP, and 3-BP are mostly excreted in urine in their glucuronidated form. If the free form is the bioactive species, the urinary distribution of free and conjugated species may provide valuable information for exposure assessment and risk assessment.