Computational Simulation On The Endocrine Disrupting Potency Mediated By Nuclear Receptor Of Some Organic Chemicals

Endocrine-disrupting chemicals (EDCs) are exogenous substances that can interfere with the function of hormonal signaling systems and elicit a range of developmental, reproductive, neurological, immune, or metabolic diseases in humans and wildlife. Distinct molecular mechanisms of EDCs were identified, including mimicking natural hormones, antagonizing hormones action or modifying hormones synthesis, metabolism, and transport. Mimicking nuclear receptors (NRs) is the major mechanism of actions for EDCs. By far the largest and best investigated endocrine disrupting effects are those for estrogen receptor (ER) Xenoestrogens with hydroxyl group can ionize as both neutral and ionic forms in aqueous solution. However, it is unclear whether the estrogenic effects of those compounds are related to their chemical forms up to now. Meanwhile, more and more EDCs were identified as potential agonists for thyroid hormone receptor (TR) and pregnane X receptor (PXR), which are also the important NR superfamily members. Nevertheless, little knowledge about the mechamisms of their agonistic activities is available. Based on the computational toxicology methods such as Quantitative Structure-Activity Relationship (QSAR) and molecular docking, we shed light on the underlying ER agonistic effect mechanisms of ionizable compounds and characterize the mechanisms of the agonistic activities of partial EDCs on TR and PXR.According to ionization equilibrium, the acidity coefficient (pKa) was used to characterize the ionization effect of model compounds (hydroxylated polychlorinated biphenyls, HO-PCBs) in pH=7.40. Based on the significant negative correlation between pKa and the ER agonistic effect of HO-PCBs (\ogRP), we theorize that the chemical forms may influence the ER agonistic effect and the HO-PCBs with ionic form has more potency than that with neutral forms. The QSAR model and molecular docking results further decipher that the penetration process of the model compounds into cell rather than binding process of the HO-PCBs with ERa is influenced by the ionization effect.Based on the QSAR development and validation guidelines issued by Organization for Economic Co-operation and Development, two QSAR models were constructed to characterize the disrupting effects of hydroxylated polybrominated diphenyl ethers (HO-PBDEs) on human TR(3agonistic effect and8sorts of pesticides on human PXR agonistic effect. Three structural descriptors (ATS8m, RDFo3u, and L2e) were selected to derive the QSAR model characterizing the hTRp agonistic effect of HO-PBDEs. This QSAR model has good goodness-of-fit (R2Y=0.94) and robustness (Q2cum=0.93, RMSE=0.30). The external validation and applicability domain (AD) evaluates result indicate that the model has good predictivity (Q2ext=0.62, RMSE=0.73) and none outliers. The developed model indicates that molecular volume and atom electronegativity are important factors governing the thyroid hormone activities of HO-PBDEs. Thus, the model proposed can be used to fast and accurately predict the hormone activity of other HO-PBDEs to TRβ.Furthermore, eight structural descriptors were introduced in the QSAR model characterizing the human PXR agonistic activities (expressed by-logREC2o) on8sorts of pesticides (organochlorines, diphenyl ethers, ureas, triazines, organophosphorus, pyrethroids, carbamates, and acid amides). Model assessment and external validation results indicate that the model has good goodness of fit (R2=0.75), robustness (R2adj=0.72), and predictive ability (Q2ext=0.51). The Euclidean distance and Williams plot results show that all chemicals both in training set and in validation set are in AD. The developed model indicates that the agonistic activities of the pesticides for human PXR are mainly influenced by molecular volume and distance between atoms, and have less relationship with molecular charges, atom electronegativity, and unsaturation. Therefore, the developed QSAR model could be used to high-throughput screen potential human PXR agonists from other pesticides with similar structures.