PhD Scientific Days 2022

Pharmaceutical Sciences II. (Poster discussion will take place in the Aula during the Coffee Break)

From Structural Dynamics to Machine Learning: SULT1 Isoenzymes substrate binding and selectivity

Text of the abstract

From Structural Dynamics to Machine Learning: SULT1 Isoenzymes substrate binding and selectivity
D. Toth# , B. Dudas, Y. Bagdad, D. Perahia, E. Balog*, M. A. Miteva*
Introduction : Sulfotransferase enzymes (SULTs) are a superfamily of cytosolic proteins involved in the metabolism of endogenous compounds and xenobioitics. By catalysing a sulfate transfer from their co-factor 3′-Phosphoadenosine 5′-Phosphosulfate (PAPS) they eliminate a large variety of small molecules like drugs, natural compounds, hormones and neurotransmitters. Even though the tertiary structure across the family is very similar, their substrates vary considerably in size and structure.
Aims: The aim of our project is to better understand the molecular mechanisms driving the selectivity between three different SULT1 isoforms, the broadest SULT1A1, the dopamine selective SULT1A3 and the estrogen selective SULT1E1.
Method: In our study at the Semmelweis University, we employed molecular dynamics (MD) simulations and the recently developed approach of MD with excited Normal Modes (MDeNM) to elucidate molecular mechanisms guiding the recognition of diverse substrates and inhibitors. At University of Paris Cite – Inserm U1268 laboratory located in the Faculty of Pharmacy we employ a virtual screening method by using docking and scoring on the generated conformational ensembles on each isoforms.
Results: MDeNM allowed exploring an extended conformational space of PAPS-bound SULT1A1, which has not been achieved up to now by using classical MD. The generated ensembles combined with docking of known SULT1A1 ligands shed new light on substrate and inhibitor binding mechanisms. Based on the achieved results on SULT1A1, we expanded our research on the other two isoform using similar molecular dynamics simulations and probing their interactions with substrates.
Conclusion: This combined method helps us to explore the substrate selectivity and further develop machine learning models capable of recognising new inhibitors and substrates specific for a particular SULT1 isoform, thus helping in the development of ADME-Tox profiling of novel drug candidates or environment xenobiotics.
Funding: TKP2021-EGA-23 Ministry of Innovation and Technology of Hungary, Development and Innovation Fund; Campus France, Ambassade de France en Hongrie, Service de Coopération et d’Action Culturelle, fichier No. 2021-0143339