Thyroid hormone signaling helps regulate many processes, including metabolism, cardiovascular function, and brain development. But manmade chemicals like herbicides and plasticizers that can disrupt this signaling have found their way into the environment. And part of that environment is pretty close to home: household dust.
Researchers at Umeå University have developed a model to help identify which of the chemicals found in household dust might be binding to the thyroid receptor and thus disrupting the signaling process.
Kwangho Nam, Patrik L. Andersson, and coworkers incorporated into their model multiple crystal structures of the β1 form of the human thyroid receptor, each complexed with a different ligand and with a slightly different conformation of the ligand binding site (Chem. Res. Toxicol. 2016, DOI: 10.1021/acs.chemrestox.6b00171). They used a set of compounds known to bind to the thyroid receptor to develop computational methods for predicting compounds that interact with the binding site. To get better predictions, they used a molecular mechanics method that better estimates how the solvent affects ligand binding.
They then used this model to screen a collection of 485 previously identified indoor dust contaminants to determine which ones, if any, might target the thyroid receptor. In that set, the model identified 31 of the compounds as possible thyroid receptor binders.
The researchers chose six of the compounds—five binders and one predicted nonbinder—to test experimentally for thyroid receptor binding. They did this using a method called isothermal titration calorimetry which measures the binding affinity of each compound with the receptor’s ligand binding domain.
Four of the five predicted binders turned out to bind weakly to the receptor, but within the range of previously reported thyroid receptor binders. Of these, the herbicide 2,4,5-trichlorophenoxyacetic acid bound most strongly. The other predicted binder—a compound known as cyclanilide, a plant growth regulator used as an herbicide—turned out to be a false positive. Information from the model could be used to prioritize chemicals for experimental testing.