The toxicokinetics of bisphenol A and its metabolites in fish elucidated by a PBTK model
Graphical abstract
Introduction
During the first decade of the 21st century, bisphenol A (BPA) has raised serious concern among the scientific community. BPA was first synthesized in 1891 and its use became widespread in the 1950s; its production exceeded six billion pounds in 2000 (Vandenberg et al., 2007). BPA was first developed as a synthetic estrogen because it can bind to estrogen receptors (ERs). However, because its binding capacity is 5 to 6 orders of magnitude smaller than that of estradiol, its use as a synthetic estrogen was abandoned (Dodds et al., 1938). Nevertheless, its use as a plastic monomer and plasticizer in various everyday products has led to ubiquitous environmental contamination. Over the last decade, BPA has been detected in freshwater courses, usually below 1 µg/L, but downstream from wastewater discharge levels reached 100 µg/L (Faheem and Bhandari, 2021). Consequently, Flint et al. (2012) determined the typical environmental BPA contamination level to be 12 µg/L or less.
Several studies have focused on the effects of BPA on aquatic communities and, in particular, its endocrine disrupting effects on reproduction and growth process in fish. Moreover, recent papers have demonstrated that exposure to BPA triggers a variety of effects that are not only related to reproduction but also innate immunity, cardiac response, and oxidative stress (Wu et al., 2011; Little and Seebacher 2015; Qiu et al., 2016; Pandey et al., 2018; Gu et al., 2020). In the context of environmental risk assessment, effects at higher levels of organization are extrapolated from individual-level data. However, extrapolations generally fail because the link between each level is mostly empirical (Forbes et al., 2008). The first step in building a mechanistic framework to assess the environmental risk of BPA is to describe the relationship between environmental levels and internal levels in the organism, i.e. the toxicokinetics. On this basis, physiologically-based toxicokinetic (PBTK) models have been proven to be valuable tools to predict how fish accumulate chemicals based on ADME (Absorption, Distribution, Metabolization, and Excretion) processes (Gerlowski and Jain 1983; Brinkmann et al., 2016; Grech et al., 2017; Tebby et al., 2019; Mit et al., 2021). PBTK models have the advantage of simulating the time-course of toxicant concentrations over time and can also provide a mechanistic framework to improve the understanding of the contribution of each ADME process.
Two PBTK models have already been used to simulate the TK of BPA in fish. Both models were generic, meaning that they were not specific to BPA but initially applicated to various substances. Pery et al. (2014) developed a PBTK for zebrafish (Danio rerio) with eight compartments, and Grech et al. (2019) extended the PBTK to three more species used in ecotoxicology, fathead minnow (Pimephales promelas), rainbow trout (Oncorhynchus mykiss), and three-spined stickleback (Gasterosteus aculeatus). The model proposed by Pery et al. (2014) was successfully evaluated on only one study reporting BPA zebrafish whole-body concentrations (Lindholst et al., 2003). The model presented by Grech et al. (2019) was evaluated on both BPA whole-body and organ concentrations in zebrafish and trout (Lindholst et al., 2000; Lindholst et al., 2001; Lindholst et al., 2003; Fang et al., 2016). However, in those models, metabolization was extremely simplified since it was treated as a way of excretion, many parameters were retrieved from publications related to mammals (Shin et al., 2004; Edginton and Ritter 2009) and, since no calibration was performed on the model parameters, ten-fold overpredictions compared to data from the literature were sometimes observed (Grech et al., 2019).
Few data are available regarding the ADME processes of BPA in fish. As in mammals, BPA undergoes phase II metabolism, which is assumed to occur mainly in the liver (Lindholst et al., 2001; Lindholst et al., 2003). Interestingly, the enzymes involved in metabolite synthesis have been characterized since zebrafish cells are commonly used as human metabolization models. The activity of UDP glucuronosyltransferases (UGTs) and sulfotransferases (STs) in the presence of BPA result in the fast production of BPA monoglucuronide (BPA gluc) and BPA monosulfate (BPA sulf) (Lindholst et al., 2001; Lindholst et al., 2003; Ohkimoto et al., 2003; Wang et al., 2014). BPA gluc was shown to be the main metabolization product in mammals and fish, with BPA sulf production being a minor pathway (Lindholst et al., 2003; Gramec Skledar and Peterlin Mašič, 2016). Thus, BPA gluc levels reached 22 times the steady-state whole body BPA concentration in zebrafish, whereas BPA sulf only represented less than a tenth of the BPA level (Lindholst et al., 2003). Nevertheless, it has been shown that BPA gluc cannot bind to ERs and cannot be considered an EDC (Matthews et al., 2001). Yet, as pointed out in Karrer et al. (2018), some mechanisms of action of BPA metabolites may still be unknown. For example, Boucher et al. (2015) showed that BPA gluc induced adipocyte differentiation when it was supposed to be inactive. In addition, BPA metabolites could be subject to deconjugation that could increase the actual concentration in fish organs, as shown in rats (Kawamoto et al., 2007).
This study aimed to propose a new PBTK explicitly developed for BPA in the three-spined stickleback. To this purpose, we modified the PBTK of Grech et al. (2019) to describe the kinetics of the two primary BPA metabolites, BPA gluc, and BPA sulf, in fish organs. New kinetic data in stickleback was obtained and used to calibrate the model. PBTK predictions were then compared to observed data found in the literature for zebrafish and trout. The inter-species variability of BPA kinetics was then studied using the PBTK model and discussed to improve the mechanistic understanding of BPA TK.
Section snippets
Literature experimental data
Five publications were identified in the literature to provide TK data for the calibration of the model (Lindholst et al., 2000; Lindholst et al., 2001; Lindholst et al., 2003; Fang et al., 2016; Eser et al., 2021). A description of each experimentation can be found in Table 1 and an extended description can be found in SI Section 7. Experiments had been carried out on rainbow trout and zebrafish with various exposure scenarios. BPA levels were measured at 1, 5, 12, and 21 days (
Determination of critical parameters
The sensitivity analysis performed prior to calibration (SI Section 3. Sensitivity analysis) showed a varying influence of parameters depending on the exposure period. In most cases, BPA concentration in organs was mainly driven by BPA-related parameters, PCbw, UF, and the plasmatic clearance describing BPA gluc synthesis, but also by plasma parameter (percentage inverse of the hematocrit). Metabolite concentrations were driven by BPA-related and metabolite-related parameters such as plasmatic
Discussion
Fish have been used as a model species for endocrine-disrupting effects for many reasons, including their ease of use in laboratory experiments, the ubiquity of endocrine chemicals in water, and also because some species, such as three-spined stickleback, have shown specific biomarkers indicative of the presence of EDCs (Tyler et al., 1998; Jolly et al., 2009). In particular, BPA effects, and to a lesser extent, BPA kinetics, were the subject of several publications (Flint et al., 2012;
Conclusion
An expected output of this work is to propose an integrating approach to link internal concentrations to BPA effects in three-spined stickleback in the future. Thus, this work consisted of the first step of a scaling-up process where environmentally relevant concentrations of BPA are linked to bioaccumulation in fish organs. The structure of the model was built to allow prediction of BPA concentrations and its metabolites in various organs. The model was then calibrated based on TK data
Funding information
This work was supported by the French National program EC2CO (Ecosphère Continentale et Côtière) as part of the DERBI project and by the French Ministry of ecological transition (P190). The authors wish to thank Cleo Tebby for its contribution in the calibration process and its careful proofreading.
CRediT authorship contribution statement
Corentin Mit: Investigation, Methodology, Software, Formal analysis, Writing – original draft, Writing – review & editing. Anne Bado-Nilles: Supervision, Writing – review & editing. Gaëlle Daniele: Formal analysis, Investigation, Writing – review & editing. Barbara Giroud: Formal analysis, Investigation, Resources, Writing – review & editing. Emmanuelle Vulliet: Formal analysis, Investigation, Resources, Writing – review & editing. Rémy Beaudouin: Formal analysis, Software, Supervision, Writing
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References (63)
- et al.
Presystemic branchial metabolism limits di-2-ethylhexyl phthalate accumulation in fish
Toxicol. Appl. Pharmacol.
(1989) - et al.
Modelling BPA effects on three-spined stickleback population dynamics in mesocosms to improve the understanding of population effects
Sci. Total Environ.
(2019) - et al.
Human exposure to bisphenol A by biomonitoring: methods, results and assessment of environmental exposures
Toxicol. Appl. Pharmacol.
(2008) - et al.
A model for exchange of organic chemicals at fish gills: flow and diffusion limitations
Aquat. Toxicol.
(1990) - et al.
Detrimental effects of bisphenol compounds on physiology and reproduction in fish: a literature review
Environ. Toxicol. Pharmacol.
(2021) - et al.
Bisphenol A exposure, effects, and policy: a wildlife perspective
J. Environ. Manag.
(2012) - et al.
Physiologically based pharmacokinetic modeling: principles and applications
J. Pharm. Sci.
(1983) - et al.
Bisphenol A and its analogs: Do their metabolites have endocrine activity?
Environ. Toxicol. Pharmacol.
(2016) - et al.
Toxicokinetic models and related tools in environmental risk assessment of chemicals
Sci. Total Environ.
(2017) - et al.
Generic physiologically-based toxicokinetic modelling for fish: integration of environmental factors and species variability
Sci. Total Environ.
(2019)