Objective
The objective of the investigation was to study placental transfer and conjugation of bisphenol A (BPA) across the human placenta.
Study Design
Human placentae obtained from healthy term singleton pregnancies were utilized in a dual recirculating model of ex vivo placental perfusion. Seven placentae were perfused with BPA (10 ng/mL) added to the maternal perfusate for 180 minutes. Antipyrine and fluorescein isothiocyanate dextran were used as positive and negative controls, respectively, to validate integrity of the circuits. Concentrations of BPA and its conjugates were determined by liquid chromatography-mass spectrometry.
Results
The transfer percentage for antipyrine and BPA were 25.5 ± 1.13% and 27.0 ± 1.88%, respectively, and the transfer index for BPA was 1.1 ± 0.09 after 180 minutes of perfusion. Only 3.2 ± 1.6% of BPA in the fetal compartment was in the conjugated form.
Conclusion
Bisphenol A at low environmentally relevant levels can transfer across the human placenta, mainly in active unconjugated form.
Bisphenol A (BPA) is a well-studied xenoestrogen with weak affinity for the estrogen receptor that is used in the preparation of polycarbonate plastics, dental sealants, and resins for can linings. Residual traces remain after manufacture and are slowly released, such that BPA in foodstuffs and drinking water is now the most significant source of exposure to synthetic xenoestrogens in New Zealand.
Recent influential publications associating BPA exposure with several diseases have led to polycarbonate plastics being banned in the production of baby bottles in Canada and to some manufacturers switching to BPA-free plastics.
The estrogenic action of BPA has been demonstrated in both in vivo and in vitro studies. BPA binds to both nuclear and membrane estrogen receptors and imparts both genomic and nongenomic steroid actions. Experiments in rodents have demonstrated that BPA can readily cross the placenta and can cause developmental abnormalities in the offspring. The transgenerational effects of BPA on the expression patterns of some testicular steroidogenic coregulators have been reported in the rat. These and several other studies have shown the endocrine-disrupting activity of BPA, especially if there is fetal and/or neonatal exposure at vulnerable periods of development.
BPA has been detected in maternal and fetal plasma, placenta, amniotic fluid, and follicular fluid. The concentrations of BPA are very variable and range from 0.3 to 18.9 ng/mL in maternal plasma, from 0.2 to 9.2 ng/mL in fetal plasma, and from 1.0 to 104.9 ng/g in term human placenta. Studies have shown that hepatic uridine 5′ diphosphate glucuronyltransferases (UGT) can conjugate BPA into its glucuronide form, which is devoid of estrogenic action. If placental UGT could also conjugate BPA, then the developmental abnormalities in the fetus caused by exposure to the potent form of BPA would be minimized.
To the best of our knowledge, none of the studies directly assessed the ability of the human placenta to conjugate BPA. We hypothesize that BPA at environmentally relevant concentration can transfer across the human placenta in an active unconjugated form. Controlled experiments in pregnant mothers are lacking because the consequences on the developing fetus of exposure to these compounds make such a study impossible in women. Placental perfusion of human placentae is the only viable alternative.
We previously validated a placental perfusion model to study the transplacental transfer of zidovudine, an anti–human immunodeficiency virus drug across the human placenta.
The aim of this study was to determine placental transfer and conjugation of BPA at environmentally relevant concentrations.
Materials and Methods
Placental perfusion
Placentae were obtained from women undergoing elective cesarean section at term from Auckland City Hospital (Auckland, New Zealand). Written consents were obtained from each pregnant woman for the use of placentae for research, and the use was approved by the regional ethics committee. Placentae from healthy nonsmoking mothers were used in the study.
A modified and fully validated dually perfused ex vivo placental perfusion system was used to study the transfer of BPA across the human placenta. The placentae were transported to the laboratory within 30 minutes of delivery. Every placenta collected for perfusion was thoroughly examined visually for gross breakages of the villous structures to try to find suitable lobules for perfusion. Placentae with visible gross breakages were excluded.
Perfusion was established in the placentae in which a suitable lobule could be found. The chorionic vessels were catheterized with polyethylene tubes, and the perfusate (phenol red-free medium-199 [M-3769; Sigma Aldrich, St Louis, MO]) cell culture media supplemented with 25 g/L polyvinylpyrrolidone PVP-40 (Sigma Aldrich), 1 g/L bovine serum albumin–fraction V (Sigma Aldrich), 2 g/L glucose (Sigma Aldrich), 20,000 IU/L heparin (Multiparin, CP Pharmaceuticals Ltd, Wexham, UK), and 48 mg/L gentamicin reagent solution (GIBCO; Invitrogen, Aukland, New Zealand) were allowed to circulate through the fetal side by means of a digitally controlled pump.
While perfusate circulated in the fetal side, the cotyledon was severed from the remaining placenta and mounted in the perfusion chamber, with the fetal side facing up. The maternal side was catheterized, and the perfusate was allowed to circulate through it with the help of a digitally controlled pump. After 15 minutes, both compartments were perfused in a recirculating mode, and the perfusion was started with a maternal flow rate of 10 mL/min and a fetal flow rate of 4 mL/min for 1 hour to equilibrate the ex vivo system. The maternal circuit was gassed with 95% O 2 and 5% CO 2 (Carbogen gas). After an hour of preperfusion (equilibration), the perfusate in the maternal compartment was replaced with fresh perfusate containing BPA (Sigma Aldrich) at 10 ng/mL, antipyrine (Sigma Aldrich) at 40 μg/mL, and fluorescein isothiocyanate (FITC)-dextran (FITC-DX, FD4 [Sigma Aldrich]) at 12.5 μg/mL.
The BPA concentration selected is within the range from the serum levels reported in pregnant women. The perfusion was then continued for 3 hours, and the samples from maternal and fetal reservoirs were collected at 30 minute intervals and stored at –80°C. Perfusions were stopped in any placentae with visible or measured leakages that were 3 mL/h or greater and were rejected from the study. Hence, every single placenta that was used to produce the data in this study has met all of our quality control measures including the following: (1) a fetal flow rate of 4 mL/min; (2) a maternal flow rate of 10 mL/min; (3) a fetal pressure 30-60 mm Hg; (4) a pH 7.2-7.4; and (5) a fetomaternal fluid shift of less than 3 mL/hr.
The viability and metabolic activity were assessed by measuring glucose utilization, lactate production (Hitachi 902 autoanalyzer; Hitachi High Technologies Corp, Tokyo, Japan), and human chorionic gonadotropin (hCG) secretion (enzyme-linked immunosorbent assay). The FITC-DX transfer was measured with a spectrofluorometer and the antipyrine transfer was measured by high-performance liquid chromatography (HPLC; ultraviolet [UV] detection). A liquid chromatography-tandem mass spectrometry (LC-MS/MS) method was optimized to measure the BPA levels.
Antipyrine detection with HPLC (UV)
All the reagents used for HPLC and LC-MS/MS were of HPLC grade (Merck, KGaA, Darmstadt, Germany). The HPLC method was a modification of the method standardized in our laboratory. Briefly 100 μL of phenacetin (internal standard; Serva Feinbiochemica, Hesperia, CA), 500 μL of 0.025 N HCl, and 500 μL of acetonitrile were added to 500 μL of perfusate and mixed well by vortexing. Then 4 mL of ethyl acetate was added, and the solution was mixed for 15 minutes in a tube rotator. The tubes were then centrifuged at 3000 × g for 10 minutes (J6-MI; Beckman) and stored at –80°C for 30 minutes. The supernatant (organic phase) was carefully transferred to a glass tube and lyophilized. The dried extracts were redissolved in 200 μL of mobile phase, mixed thoroughly by vortexing, and transferred to 200 μL polyvinylchloride inserts in HPLC vials and sealed with thin septa and screw caps.
Chromatographic separations were performed on an HPLC system (Waters Alliance 2690; Waters Corp, Milford, MA). The column used to separate the compounds was Luna 3μ C18 (2) 100A, 250 × 4.6 mm (Phenomenex, Torrance, CA). The mobile phase was 6.7 mM phosphate buffer (pH 7.2)–acetonitrile (65:35). The separation was carried out isocratically at a flow rate of 0.7 mL/min. Ten microliters of sample was injected and detection was carried out using a UV detector (996 photodiode array detector; Waters) at 254 nm. The data were analyzed using Waters Millennium Chromatography Manager (version 4).
BPA detection with LC-MS/MS
An LC-MS/MS method was optimized in our laboratory to measure the BPA content in the perfusates. All the reagents used for analysis were filtered through Empore extraction disk (RP-SDB discs; Phenomenex, Torrance, CA) to prevent any potential BPA contamination. Briefly 100 μL of a mixture of BPA d 16 (30 ng/mL) (deuterium-labeled internal standard for BPA; Cambridge Isotope Laboratory, Woburn, MA) was mixed in a tube with 200 μL of perfusate, and 1 mL of ethyl acetate was then added and the solution mixed by vortexing for 30 seconds. The tubes were then kept in a tube rack for 10 minutes.
The organic phase was collected in a collection glass tube and lyophilized. The extract was redissolved in 70 μL of mobile phase (methanol-water, 80:20), vortexed, and transferred to 200 μL PVC inserts in HPLC vials. Ten microliters of the samples were injected with an autoinjector. The column used for separating the compounds was a Phenomenex 100 × 3.0 mm 2.5μ C18 (2) HST. The HPLC system (Alliance 2690; Waters) was interfaced to a triple quadruple mass spectrometer (Finnigan TSQ Quantum Ultra AM; Thermo Electron Corp, San Jose, CA). The parent-to-daughter ion transition for BPA was 227.03-133.03 and for BPA d 16 was 241.17-142.16. The retention time was 6.47 minutes for BPA and 6.39 minutes for BPA d 16 . The data were analyzed using Finnigan Excalibur software (Thermo Electron Corp).
To detect the conjugated levels of BPA, a deconjugation enzyme assay was used with modification. Two hundred microliters of perfusate was buffered with 50 μL of 2 N sodium acetate (pH 5.0). Glucuronidase/sulfatase (Crude extract from Helix pomatia ; Sigma Aldrich) at preoptimized concentration (500 U of glucuronidase and 3.5 U of sulfatase) was added to tubes and incubated at 37°C in a water bath overnight (14 hours). The reaction was stopped by heating the tubes at 70°C for 10 minutes. The total BPA was then extracted as per the protocol cited in previous text, and detection was carried out using LC-MS/MS as described in previous text. The BPA detected before deconjugation (free BPA) was deducted from BPA detected after deconjugation (total BPA), and the difference (conjugated BPA) is expressed as a percent of total BPA.
Calculations
The formula used to calculate transfer percentage from maternal to fetal circulation was as follows :
Transfer percentage = 100 × Fc X Fv/ [(Fc X Fv) + (Mc X Mv)], in which Fc is fetal concentration, Fv is fetal volume, Mc is maternal concentration, and Mv is maternal volume.
The transfer index = transfer percentage of BPA/transfer percentage of antipyrine. Values are expressed as mean ± SEM.
Results
Placental perfusion
A reduction in fetal volume and a consequent surplus in maternal volume were observed whenever there was a leak in the system. All placentae selected for the study showed a fetomaternal fluid loss of less than 3 mL/h for a total of 180 minutes of perfusion. The recorded fetal pressure for all successful placentae was in the range of 25-40 mm Hg. None of the placentae showed fluctuations in fetal pressures once they had passed the equilibration period. The wet weight of the perfused cotyledon was 29.3 ± 2.95 g.
The placentae studied maintained viability as well as metabolic activity throughout the perfusion period as evident from constant glucose consumption, lactate production, and β-hCG secretion. The values for glucose consumption and lactate production were 0.32 ± 0.06 μM/g per minute and 0.52 ± 0.1 μM/g per minute in the maternal compartment and 0.32 ± 0.06 μM/g per minute and 0.35 ± 0.06 μM/g per minute in the fetal compartment. β-hCG secretion was observed only in the maternal compartment (3.0 ± 0.9 mIU/g per minute), and the values in the fetal compartment were below the detection level.
Antipyrine and FITC-DX detection
HPLC-UV detection was used to measure antipyrine concentrations in maternal and fetal compartments. Antipyrine transfer from maternal to fetal compartment was evident from the gradual disappearance of antipyrine in the maternal compartment and subsequent gradual appearance of antipyrine in the fetal compartment. Approximately 25% of initial maternal concentration of antipyrine was found in the fetal compartment after 180 minutes of perfusion ( Figure 1 , A). FITC-dextran transfer was negligible after 180 minutes of perfusion (<1% per hour) ( Figure 1 , B).
