Antipsychotic Haloperidol Binding to the Human Dopamine D3 Receptor: Going beyond Docking Through QM/MM Re- finement for the Design of Improved Schizophrenia Medicines Geancarlo Zanatta a *, Gustavo Nunes a , Eveline M. Bezerra b , Roner F. da Costa c , Alice Martins b , Ewerton W. S. Caetano d , Valder N. Freire e , Carmem Gottfried a a Department of Biochemistry at Federal University of Rio Grande do Sul, 90035-003 Porto Alegre, RS, Brazil; b Post-graduate Program in Pharmaceutical Sciences, Pharmacy Faculty, Federal University of Ceará, 60430-372 Fortaleza, CE, Brazil; c Department of Physics, Universidade Federal Rural do Semi-Árido, 59780-000 Caraúbas, RN, Brazil; d Federal Institute of Education, Science and Technology, 60040-531 Fortaleza, CE, Brazil; e Department of Physics at Federal University of Ceará, 60455-760 Fortaleza, CE, Brazil. KEYWORDS: Quantum biochemistry; Dopamine receptor; Haloperidol; D3 binding pocket; Quantum mechanics; DFT; antipsychotic; ab initio; docking; QM/MM ONIOM method ABSTRACT: As the dopamine D3R receptor is a promising target for schizophrenia treatment, an improved understanding of the binding of existing antipsychotics to this receptor is crucial for the development of new potent and more selective therapeutic agents. In this work, we have used X-ray co-crystallization data of the antagonist eticlopride bound to D3R as a template to predict, through docking essays, the placement of the typical antipsychotic drug haloperidol at the D3R receptor binding site. Afterwards, classical and quantum mechanics/molecular mechanics (QM/MM) computations were employed to improve the quality of the dock- ing calculations, the QM part of the simulations being accomplished by using the density functional theory (DFT) formalism. After docking, the calculated QM improved total interaction energy E QMDI = -170.1 kcal/mol was larger (in absolute value) than that obtained with classical molecular mechanics improved (E CLDI = -156.3 kcal/mol) and crude docking (E CRDI = -137.6 kcal/mol) pro- cedures. The QM/MM computations reveal the pivotal role of the Asp110 amino acid residue in the D3R haloperidol binding, fol- lowed by Tyr365, Phe345, Ile183, Phe346, Tyr373 and Cys114. Besides, it highlights the relevance of the haloperidol hydroxyl group axial orientation, which interacts with the Tyr365 and Thr369 residues enhancing its binding to dopamine receptors. Finally, our computations indicate that functional substitutions in the 4-clorophenyl and in the 4-hydroxypiperidin-1-yl fragments (such as C3H and C12H hydrogen replacement by OH, OOH or Cl groups) can lead to haloperidol derivatives with distinct dopamine an- tagonism profiles. The results of our work are a first step towards in silico quantum biochemical design and probing of new medi- cines to treat schizophrenia.