Binding interaction of typical emerging contaminants on Gobiocypris rarus transthyretin: an in vitro and in silico study

Abstract

Emerging contaminants (ECs) have drawn global concern, and the endocrine disrupting chemicals is one of the highly interested ECs categories. However, numerous ECs lacks the basic information about whether they can disturb the endocrine related biomacromolecules or elicit endocrine related detrimental effects on organism. In this study, the potential binding affinity and underlying binding mechanism between 29 ECs from 7 chemical groups and Gobiocypris rarus transthyretin (CrmTTR) are investigated and probed using in vitro and in silico methods. The experimental results demonstrate that 14 selected ECs (11 disinfection byproducts, 1 pharmaceuticals and personal care product, 1 alkylphenol, 1 perfluoroalkyl and polyfluoroalkyl substance) are potential CrmTTR binders. The CrmTTR binding affinity of three ECs (i.e., 2,6-diiodo-4-nitrophenol (logRP(T₄) = 0.678 ± 0.198), 2-bromo-6-chloro-4-nitrophenol (logRP(T₄) = 0.399 ± 0.0908), tetrachloro-1,4-benzoquinone (logRP(T₄) = 0.272 ± 0.0655)) were higher than that of 3,3′,5,5′-tetraiodo-L-thyronine, highlighting that more work should be performed to reveal their potential endocrine related harmful effects on Gobiocypris rarus. Molecular docking results imply that hydrogen bond and hydrophobic interactions are the dominated non-covalent interactions between the active disruptors and CrmTTR. The optimum mechanism-based (for CrmTTR), and high throughput screening (for CrmTTR, little skate-TTR, seabream-TTR, and human-TTR) binary classification models are developed using three machine learning algorithms, and all the models have good classification performance. To facilitate the use of developed high throughput screening models, a tool named “TTR Profiler” is derived, which could be employed to determine whether a given substance is a potential CrmTTR, little skate-TTR, seabream-TTR, or human-TTR disruptor or not.