In vivo and in silico assessments of estrogenic potencies of bisphenol A and its analogs in zebrafish (Danio rerio): Validity of in silico approaches to predict in vivo effects

This study assessed the estrogen-like potencies of bisphenol A (BPA) and its analogs (BPs) using in vivo and in silico approaches in zebrafish. Zebrafish embryos were exposed to 16 BPs, most of which concentration-dependently induced cytochrome P450 19A1b (CYP19A1b) expression. BPs-induced CYP19A1b expression was suppressed by fulvestrant, a nonselective high affinity antagonist for estrogen receptor (Esr) subtypes. For BPs that concentration-dependently induced CYP19A1b expression, we estimated their 50 % effective concentration (EC50) and relative potencies (REPs) with respect to the potency of BPA for inducing CYP19A1b expression. BP C2, Bis-MP, and BPAF showed lower EC50 than BPA, BPE, and BPF, while BPZ and BPB showed moderate EC50. The REP order of the BPs was BP C2 (26) > Bis-MP (24) > BPAF (21) > BPZ (5.8) > BPB (2.7) > BPE (1.5) > BPF (0.63) > 2,4′-BPF (0.22), indicating that some BPs showed greater estrogenic potencies than BPA in our system. We also constructed in silico homology models of ligand binding domains for zebrafish Esr subtypes, including Esr1, Esr2a, and Esr2b. Molecular docking simulations of ligands with the Esr subtypes revealed the interaction energies of some BPs were lower than that of BPA. The interaction energies showed significant positive correlations with their EC50 values for inducing CYP19A1b expression in vivo. This study showed that some BPA analogs have greater estrogenic potencies than BPA and that in silico simulations of interactions between ligands and Esr subtypes may help predict in vivo estrogenic potencies of untested chemicals. [Display omitted] •Zebrafish CYP19A1b is a good marker to assess estrogenic potencies of bisphenols.•Some BPA alternatives showed greater CYP19A1b induction potencies than BPA.•Homology models of zebrafish Esr subtypes were constructed.•In silico BPs-Esr docking simulation may help predict in vivo estrogenic potencies.