© 2017 American Chemical Society. This study presents an implementation of the protein-ligand docking program GOLD and a generalizable method to predict the binding site and orientation of potential vanadium drugs. Particularly, theoretical methods were applied to the study of the interaction of two VIVO complexes with antidiabetic activity, [VIVO(pic)2(H2O)] and [VIVO(ma)2(H2O)], where pic is picolinate and ma is maltolate, with lysozyme (Lyz) for which electron paramagnetic resonance spectroscopy suggests the binding of the moieties VO(pic)2 and VO(ma)2 through a carboxylate group of an amino acid residue (Asp or Glu). The work is divided in three parts: (1) the generation of a new series of parameters in GOLD program for vanadium compounds and the validation of the method on five X-ray structures of VIVO and VV species bound to proteins; (2) the prediction of the binding site and enantiomeric preference of [VO(pic)2(H2O)] to lysozyme, for which the X-ray diffraction analysis displays the interaction of a unique isomer (i.e., OC-6-23-Δ) through Asp52 residue, and the subsequent refinement of the results with quantum mechanics/molecular mechanics methods; (3) the application of the same approach to the interaction of [VO(ma)2(H2O)] with lysozyme. The results show that convenient implementation of protein-ligand docking programs allows for satisfactorily reproducing X-ray structures of metal complexes that interact with only one coordination site with proteins and predicting with blind procedures relevant low-energy binding modes. The results also demonstrate that the combination of docking methods with spectroscopic data could represent a new tool to predict (metal complex)-protein interactions and have a general applicability in this field, including for paramagnetic species.
|Publication status||Published - 6 Nov 2017|