In Silico Mechanistic Profiling to Probe Small Molecule Binding to Sulfotransferases

Abstract

Drug metabolizing enzymes play a key role in the metabolism, elimination and detoxification of xenobiotics, drugs and endogenous molecules. While their principal role is to detoxify organisms by modifying compounds, such as pollutants or drugs, for a rapid excretion, in some cases they render their substrates more toxic thereby inducing severe side effects and adverse drug reactions, or their inhibition can lead to drug–drug interactions. We focus on sulfotransferases (SULTs), a family of phase II metabolizing enzymes, acting on a large number of drugs and hormones and showing important structural flexibility. Here we report a novel in silico structure-based approach to probe ligand binding to SULTs. We explored the flexibility of SULTs by molecular dynamics (MD) simulations in order to identify the most suitable multiple receptor conformations for ligand binding prediction. Then, we employed structure-based docking-scoring approach to predict ligand binding and finally we combined the predicted interaction energies by using a QSAR methodology. The results showed that our protocol successfully prioritizes potent binders for the studied here SULT1 isoforms, and give new insights on specific molecular mechanisms for diverse ligands’ binding related to their binding sites plasticity. Our best QSAR models, introducing predicted protein-ligand interaction energy by using docking, showed accuracy of 67.28%, 78.00% and 75.46%, for the isoforms SULT1A1, SULT1A3 and SULT1E1, respectively. To the best of our knowledge our protocol is the first in silico structure-based approach consisting of a protein-ligand interaction analysis at atomic level that considers both ligand and enzyme flexibility, along with a QSAR approach, to identify small molecules that can interact with II phase dug metabolizing enzymes.

In Silico Prediction of Human Sulfotransferase 1E1 Activity Guided by Pharmacophores from Molecular Dynamics Simulations

“…To date, only two in silico prediction models have been reported that utilize a structure-based approach including protein flexibility (28,63) and one of them proposed activity prediction for SULT subtype 1E1 (28). Both employ the strategy of performing MD simulations to sample the conformational space of different SULT subtypes as a basis for their prediction model.…”

“…Based on the enzyme structure, computational studies have investigated the mechanism of inhibition of SULT1E1 by nucleotides (25) and stereoselectivity of sulfonation via docking (26), analyzed the sulfonation reaction using QM/MM methods (27), and utilized molecular docking to predict ligand binding (28). Here, we report on a novel approach of computer-based metabolism prediction for human SULT1E1.…”

Computational Biology and Chemistry in MTi: Emphasis on the Prediction of Some ADMET Properties

“…Following a similar strategy, we expanded these studies to other SULT1 isoforms (Figure ). Our best models, introducing predicted protein‐ligand interaction energy by using molecular dynamics, docking and machine learning methodologies, showed accuracy of 67.28 %, 78.00 % and 75.46 %, for the isoforms SULT1 A1, SULT1 A3 and SULT1E1, respectively (accuracy here means the correct fraction or the percentage of correctly predicted inhibitors and non‐inhibitors). To the best of our knowledge, our protocol is the first in silico structure‐based approach consisting of a protein‐ligand interaction analysis at atomic level that considered both ligand structures and enzyme flexibility, along with a QSAR approach, to identify small molecules that can interact with phase II drug metabolizing enzymes.…”

Sampling of Conformational Ensemble for Virtual Screening Using Molecular Dynamics Simulations and Normal Mode Analysis

“…Recently, we have observed [44] that despite of very high druggability score of some holo Xray structures, the obtained enrichments are not always satisfactory. The druggability score is a useful evaluation but it might be not sufficient for a final selection of the best receptor conformations.…”

“…Another MD-based approach, Limoc, aims at sampling RCE appropriate for accommodating ligands, which are chemically and structurally diverse and thus unbiased toward a particular class of ligands [43]. Recently we performed successful ligand profiling of drug metabolizing enzymes sulfotransferases by employing docking to RCE generated by MD simulations combined with hierarchical conformational clustering of different binding site conformations [44].…”