Physiologically Based Pharmacokinetic (PBPK) Modeling of the Bisphenols BPA, BPS, BPF, and BPAF with New Experimental Metabolic Parameters: Comparing the Pharmacokinetic Behavior of BPA with Its Substitutes

Background: The endocrine disrupting chemical bisphenol A (BPA) has been facing stricter regulations in recent years. BPA analogs, such as the bisphenols S, F, and AF (BPS, BPF, and BPAF) are increasingly used as replacement chemicals, although they were found to exert estrogenic effects similar to those of BPA. Research has shown that only the parent compounds have affinity to the estrogen receptors, suggesting that the pharmacokinetic behavior of bisphenols (BPs) can influence their potency. Objectives: Our goal was to compare the pharmacokinetic behaviors of BPA, BPS, BPF, and BPAF for different age groups after environmentally relevant external exposures by taking into account substance-specific metabolism kinetics and partitioning behavior. This comparison allowed us to investigate the consequences of replacing BPA with other BPs. Methods: We readjusted a physiologically based pharmacokinetic (PBPK) model for peroral exposure to BPA and extended it to include dermal exposure. We experimentally assessed hepatic and intestinal glucuronidation kinetics of BPS, BPF, and BPAF to parametrize the model for these BPs and calibrated the BPS model with a biomonitoring study. We used the PBPK models to compare resulting internal exposures and focused on females of childbearing age in a two-dimensional Monte Carlo uncertainty analysis. Results: Within environmentally relevant concentration ranges, BPAF and BPS were glucuronized at highest and lowest rates, respectively, in the intestine and the liver. The predominant routes of BPS and BPAF exposure were peroral and dermal exposure, respectively. The calibration of the BPS model with measured concentrations showed that enterohepatic recirculation may be important. Assuming equal external exposures, BPS exposure led to the highest internal concentrations of unconjugated BPs. Conclusions: Our data suggest that the replacement of BPA with structural analogs may not lower the risk for endocrine disruption. Exposure to both BPS and BPAF might be more critical than BPA exposure, if their respective estrogenic potencies are taken into account. https://doi.org/10.1289/EHP2739

. Age-group specific physiological model parameters used as input for the basic PBPK models. Table S2. Published measurements of metabolism parameters for bisphenol A. Table S3. Comparison between BPA tissue/serum partition coefficients determined experimentally (Doerge et al. (2011), highlighted in grey) and with different QSARs for QSAR selection. Table S4. Observed changes of RSS and Cmax of the concentration-time curve of unconjugated bisphenol A after decreasing and increasing the values of the individual tissue/serum partition coefficients by 10% and 50%. Table S5. Qualitative evaluation and ordinal scaling of uncertainty in the PBPK model. Model parameters were classified into five different categories: Low uncertainty (L), low to medium uncertainty (LM), medium uncertainty (M), medium to high uncertainty (MH), and high uncertainty (H) (EFSA Scientific Committee 2016). Table S6. Trapezoidal distributions used to describe uncertainty in the outer loop of the 2D-MC analysis. A description of how parameters were obtained and respective references can be found in the text. Table S7. Parametrizations for truncated normal distributions used to describe variability in the 2D-MC analysis of BPA (around central values from the basic model for females of childbearing age). A description of how parameters were obtained and respective references can be found in the text. Table S8. Tissue/serum partition coefficients for BPS, BPF, and BPAF calculated with the quantitative structure-activity relationships by DeJongh et al. (1997) and Schmitt (2008), partially used as boundaries in the uncertainty distributions. Table S9. Scenario specific exposure parameters for the comparison with Hormann et al. (2014). Table S10. PBPK model parameters for bisphenol S before and after the calibration. Figure S1. Measured and modeled serum concentration-time profiles of BPA, BPA-g, and BPA-s after peroral dosing with 100 μg BPA/kg bw. Individual measurements (open circles) represent observed serum concentrations (average ± standard deviation) of 14 adults (Thayer et al. 2015). Concentration profiles for the respective volunteers (grey solid lines) were modeled using (A) the published model by Yang et al. (2015) and (B and C) adjusted models with partly different parametrizations (see Tables 7 and 8) assuming either (B) no EHR or (C) a BPA EHR rate of 10% (see Table 5 for uptake parameters). Grey solid lines in the latter two columns depict the model results with varying parameter sets, for the individual with the median BPA concentration-time profile for better clarity (for evaluating the effects of different parameter sets all individuals were considered). The sets describing the biomonitoring data best are highlighted in blue. Abbreviations: BPA, bisphenol A; bw, bodyweight; conc., concentration; EHR, enterohepatic recirculation; g, glucuronide; s, sulfate. Figure S2. Eadie Hofstee plots of enzyme kinetics of BPS, BPF, and BPAF with human liver and intestinal microsomes. Shown are averages (black circles) and ranges from minimal to maximal reaction velocities (whiskers). Abbreviations: BPAF, bisphenol AF; BPF, bisphenol F; BPS, bisphenol S; csubstrate, substrate concentration; v, reaction velocity. Figure S3. Measured and modeled serum concentration-time profiles of BPS and BPS-g after peroral dosing with 8.75 μg BPS/kg bw. Individual measurements (black circles) represent observed serum concentrations (average ± standard deviation) of 7 adults (Oh et al. 2018). Concentration profiles for the respective volunteers (grey solid lines) were modeled using (A) the adjusted PBPK model for BPA further adjusted with BPS-specific metabolism parameters (Table  7) and (B and C) the BPS-specific model calibrated to assume a higher peroral uptake and increased clearance rates of BPS and BPS-g assuming either (B) no EHR or (C) a BPS EHR rate of 67% (see Table 5 for uptake parameters). Abbreviations: BPA, bisphenol A; BPS, bisphenol S; bw, bodyweight; EHR, enterohepatic recirculation; g, glucuronide. Figure S4. Modeled concentration profiles of unconjugated BPS obtained with the basic PBPK model in serum (A) and gonads (B) for infants (6 days-3 months), toddlers (1-3 years), children (3-10 years), adolescents (10-18 years), and adults (18-45 years) after 500 ng/kg bw single peroral and dermal exposures (t=0) respectively (rough average of peroral and dermal high BPA exposure estimates for adults by the EFSA CEF Panel (2015)), see Table 5 for uptake parameters. Females are represented by solid lines, males are represented by dotted lines. Abbreviations: BPS, bisphenol S; bw, bodyweight; c, concentration; PBPK, physiologically based pharmacokinetic. Figure S5. Modeled concentration profiles of unconjugated BPA obtained with the basic PBPK model in serum (A) and gonads (B) for infants (6 days-3 months), toddlers (1-3 years), children (3-10 years), adolescents (10-18 years), and adults (18-45 years) after 500 ng/kg bw single peroral and dermal exposures (t=0) respectively (rough average of peroral and dermal high BPA exposure estimates for adults by the EFSA CEF Panel (2015)), see Table 5 for uptake parameters. Females are represented by solid lines, males are represented by dotted lines. Abbreviations: BPA, bisphenol A; bw, bodyweight; c, concentration; PBPK, physiologically based pharmacokinetic. Figure S6. Modeled concentration profiles of unconjugated BPF obtained with the basic PBPK model in serum (A) and gonads (B) for infants (6 days-3 months), toddlers (1-3 years), children (3-10 years), adolescents (10-18 years), and adults (18-45 years) after 500 ng/kg bw single peroral and dermal exposures (t=0) respectively (rough average of peroral and dermal high BPA exposure estimates for adults by the EFSA CEF Panel (2015)), see Table 5 for uptake parameters. Females are represented by solid lines, males are represented by dotted lines. Abbreviations: BPF, bisphenol F; bw, bodyweight; c, concentration; PBPK, physiologically based pharmacokinetic. Figure S7. Modeled concentration profiles of unconjugated BPAF obtained with the basic PBPK model in serum (A) and gonads (B) for infants (6 days-3 months), toddlers (1-3 years), children (3-10 years), adolescents (10-18 years), and adults (18-45 years) after 500 ng/kg bw single peroral and dermal exposures (t=0) respectively (rough average of peroral and dermal high BPA exposure estimates for adults by the EFSA CEF Panel (2015)), see Table 5 for uptake parameters. Females are represented by solid lines, males are represented by dotted lines. Abbreviations: BPAF, bisphenol AF; bw, bodyweight; c, concentration; PBPK, physiologically based pharmacokinetic. Figure S8. Measurements of Hormann et al. (2014) (symbols) against modeled individual serum profiles (solid lines) of unconjugated BPA, BPA-g, and BPA-s in three volunteers. They handled receipt paper for 4 min with both hands which were wetted with skin sanitizer and ate 10 French fries with a contaminated hand afterwards during 4 min. One hand was then cleaned and the other hand stayed contaminated until the end of blood collection (90 min in total, see Table S9 for study-specific parameters). Abbreviations: BPA, bisphenol A; g, glucuronide; s, sulfate. Table C1. Input table "Probanden" used in the basic PBPK model code for BPA.   Edginton et al. (2006). b Indicates values for female/male. c Perfusion lower than 0.1 mL/min/g tissue: muscle and skeleton (Edginton et al. 2006;ICRP 2002). d Perfusion higher than 0.1 mL/min/g tissue: heart, kidneys, small and large intestine, pancreas, spleen, and stomach (Edginton et al. 2006;ICRP 2002  The following scaling factors were applied: 32 mg microsomal protein/g liver and 99 x 106 cells/ g liver (Barter et al. 2007). b The Km was derived from an n=1; therefore, no SD was calculated. c The arithmetic mean of female and male kinetics was used in the comparison.

Input tables
Perfusion lower than 0.1 mL/min/g tissue: muscle and skeleton (Edginton et al. 2006;ICRP 2002). b Perfusion higher than 0.1 mL/min/g tissue: heart, kidneys, small and large intestine, pancreas, spleen, and stomach (Edginton et al. 2006;ICRP 2002). Abbreviations: Cmax, maximal concentration; rich, richly perfused tissue; RSS, residual sum of squares; slow, slowly perfused tissue. Physiological model parameters have been evaluated in several studies with human volunteers/patients, so that the central tendencies are well-known. Therefore, the uncertainty around their parameter values is rather small in comparison to the interindividual variability in physiology. Among these parameters, the uncertainty varies depending on whether invasive measurement techniques are needed. For example, the uncertainty is lower for the height than for the tissue volumes, as height can be measured externally so that more measurement values exist. For BPA, partitioning was investigated in an animal experiment (Doerge et al. 2011). For the other analogues, QSARs needed to be used to derive PTS. Depending on the QSAR applied, different results can be obtained. It is uncertain which QSAR reflects the situation best. The experiments investigating metabolism kinetics were conducted in vitro. The experimental conditions may not have covered all processes that are relevant in vivo. In addition, we observed a large variation of reported parameter values for the hepatic and gut glucuronidation of BPA, but cannot depict the study that represents real circumstances best. Therefore, there is a high uncertainty concerning glucuronidation kinetics of BPA, which can be quantified. For metabolism parameters for which only one study exists, the uncertainty is not necessarily smaller. Differences between BPS kinetic parametrizations before and after calibration can be used to estimate the magnitude of uncertainty for the analogues for which we could not calibrate the models. Several studies investigated the microsomal protein content in the liver and the small intestine. The range of observations is rather narrow for the liver, meaning that the concentration is easy to analyze and/or that it doesn't vary substantially. The range is much larger for the small intestine. This means that the concentration is difficult to determine and/or that there is a large inter-individual variability. Uncertainty should therefore be evaluated for the LM H H MH-H enzyme concentration in the small intestine. For consistency reasons, we also investigated the uncertainty of the hepatic enzyme concentration.

EHR
The pathway of EHR has been observed for molecules with H molecular weights (MW) higher than 500 g/mol (Roberts et al. 2002). The MW of bisphenol glucuronides ranges from 376 (BPFg) to 512 (BPAF-g) g/mol. This means that the probability of EHR taking place could depend on the respective analogue. A comparison of possible PBPK model outputs for BPA (MW of BPA-g: 404 g/mol) with the biomonitoring data by Thayer et al. (2015) showed that BPA equally could or could not undergo EHR.
The results of the biomonitoring study by Oh et al. (2018) suggest that EHR plays an important role for BPS.

Dermal absorption (fraction)
Half-life of dermal penetration

Peroral absorption (fraction)
Uptake of BPs and metabolites from gut to liver

Urinary excretion of BPs and metabolites
Several studies investigated the dermal absorption of BPA, but MH different study designs and solvents were used. In total, reported dermal absorption ranged from 9.3% to 60%. However, the range diminishes if different solvents and study designs are differentiated.
The half-life of dermal penetration varies substantially depending MH on the solvent used in the experiment. As only few studies investigated this parameter, there is significant uncertainty. Again, the range of half-lives reported diminishes if different solvents are regarded separately. The peroral absorption fraction has been derived from recoveries of LM biomonitoring studies. The two studies available (Thayer et al. 2015;Völkel et al. 2002) report recoveries of 84-109% and 118 ± 21% respectively, indicating complete or nearly complete peroral absorption. The small intestinal transit time has been characterized in humans. M-For the metabolites, only the direct transition from enterocytes to MH the liver needs to be regarded. This has been done with optimizations within the models. The parameter is more uncertain for BPF and BPAF, for which we could not calibrate the models. The clearance rates have been characterized in biomonitoring Mstudies of BPA and it has been found that the clearance rate of BPA MH resembles the creatinine clearance of a healthy adult. The individual excretion terms have been further adjusted within the model for BPA and BPS. For BPF and BPAF, we could not calibrate the excretion terms and therefore their parametrization is more uncertain. For the age a uniform distribution was used spanning from 18-45 years, the bodyweight was calculated as (height 2 ) * BMI. The volume of distribution was set equal to the plasma volume. a Perfusion lower than 0.1 mL/min/g tissue: muscle and skeleton (Edginton et al. 2006;ICRP 2002). b Perfusion higher than 0.1 mL/min/g tissue: heart, kidneys, small and large intestine, pancreas, spleen, and stomach (Edginton et al. 2006;ICRP 2002). Perfusion higher than 0.1 mL/min/g tissue: heart, kidneys, small and large intestine, pancreas, spleen, and stomach (Edginton et al. 2006;ICRP 2002).