In silico bioavailability triggers applied to direct and indirect thyroid hormone disruptors.

Area of Science:

• Environmental Toxicology
• Medicinal Chemistry
• Computational Toxicology

Background:

• Endocrine disruption, particularly interference with thyroid hormone (TH) regulation, is a growing environmental and health concern.
• Chemicals possess structural features influencing both their potential for TH disruption and their bioavailability, which modulates toxicological impact.

Purpose of the Study:

• To develop a screening-level quantitative structure-activity relationship (QSAR) model based on substructures.
• To differentiate bioavailable thyroid hormone disruptors from non-bioavailable ones.
• To encompass direct and indirect (retinoid- and AhR-mediated) modes of action.

Main Methods:

• Derived a QSAR model using literature data from 1642 compounds.
• Incorporated Lipinski's rule-of-five and the brain/blood partition coefficient (K_brain/blood).
• Validated the model using an external test set of 145 substances.

Main Results:

• The model achieved high accuracy for 1787 compounds within its application domain, with only one false negative.
• Identified key molecular descriptors (molecular weight, H-bond donors/acceptors, log K_ow) and physicochemical properties influencing bioavailability and K_brain/blood.

Conclusions:

• The developed QSAR model effectively screens for bioavailable thyroid hormone disruptors.
• The model can be integrated into decision support systems for predictive chemical safety assessment.
• It aids in evaluating the potential for both direct and indirect disruption of thyroid hormone function.