Send Orders for Reprints to reprints@benthamscience.ae Current Computer-Aided Drug Design, 2014, 10, 315-326 315 Docking Modes of BB-3497 into the PDF Active Site – A Comparison of the Pure MM and QM/MM Based Docking Strategies Tripti Kumari, Upasana Issar and Rita Kakkar * Computational Chemistry Group, Department of Chemistry, University of Delhi, Delhi-110007, India Abstract: Peptide deformylase (PDF) has emerged as an important antibacterial drug target. Considerable effort is being directed toward developing peptidic and non-peptidic inhibitors for this metalloprotein. In this work, the known peptidic inhibitor BB-3497 and its various ionization and tautomeric states are evaluated for their inhibition efficiency against PDF using a molecular mechanics (MM) approach as well as a mixed quantum mechanics/molecular mechanics (QM/MM) approach, with an aim to understand the interactions in the binding site. The evaluated Gibbs energies of binding with the mixed QM/MM approach are shown to have the best predictive power. The experimental pose is found to have the most negative Gibbs energy of binding, and also the smallest strain energy. A quantum mechanical evaluation of the active site reveals the requirement of strong chelation by the ligand with the metal ion. The investigated ligand chelates the metal ion through the two oxygens of its reverse hydroxamate moiety, particularly the N-O - oxygen, forming strong covalent bonds with the metal ion, which is penta-coordinated. In the uninhibited state, the metal ion is tetrahedrally coordinated, and hence chelation with the inhibitor is associated with an increase of the metal ion coordination. Thus, the strong binding of the ligand at the binding site is accounted for. Keywords: BB-3497 inhibitor, density functional theory, ESP charge analysis, molecular docking, MM, peptide deformylase (PDF), QM/MM. 1. INTRODUCTION Peptide Deformylase (PDF) (EC 3.5.1.31) [1-4], an essential bacterial metalloprotease, has emerged as one of the most promising and unexploited targets of antibacterial chemotherapy, as there are no clinically used drugs that target it. With cases of antibiotic resistant bacterial strains being reported more often these days, there is an immediate need to address the issue of bacterial resistance. PDF, owing to its conserved active site across the majority of bacterial strains [5, 6], and on account of its sound selectivity towards bacteria [7, 8], appears as a novel target for developing antibiotics. Several peptidic and non-peptidic inhibitors of PDF are known and have been studied experimentally by various research groups [4, 9, 10]. A few also progressed up to Phase I of clinical trials [11-16]; however, only one, GSK1322322, could proceed further to Phase II clinical development for the treatment of lower respiratory tract and skin infections [17, 18]. The realization that potent PDF inhibitors would be of significant help to the drug discovery program gave us the drive to proceed in the direction of studying PDF inhibition. With the help of molecular modeling and in silico studies, we plan to design new leads that can inhibit the enzyme considerably and also display good pharmacological properties. However, before getting into the evaluation of different inhibitors as candidate drugs, we need to understand the active site of PDF and then explore the *Address correspondence to this author at the Computational Chemistry Group, Department of Chemistry, University of Delhi, Delhi-110007, India; Tel: +911127666313; Fax: +911127665555; E-mail: rkakkar@chemistry.du.ac.in interactions that provide considerable binding affinity and inhibition. In this paper, we have therefore evaluated a peptidic inhibitor BB-3497 [19] for its affinity and inhibition by performing its molecular docking studies. We here present a comparison of the purely molecular mechanics (MM) and quantum mechanics/molecular mechanics (QM/MM) based docking strategies. 2. COMPUTATIONAL DETAILS The starting coordinates of the PDF-BB3497 complex [PDB ID: 1G27] were imported from the Research Collaboratory for Structural Bioinformatics (RCSB) Protein Data Bank and refined for our further work. The bond orders of the residues were adjusted, and hydrogen atoms added to the structure. All crystal waters beyond 5 Å from the binding ligands were deleted, and the remaining water orientations were sampled exhaustively. The active site of this protein contains the Ni 2+ ion coordinated to the N’s of the side chain of His-132 and His-136, and to the S - of the side chain of Cys-90 of the protein, and also to the –N-O - and –C=O of the reverse hydroxamate group of the inhibitor BB-3497 (Fig. 1). Before carrying out the calculations, it was important to decide the overall charge on the amino acids that are directly bonded to the metal ion in the protein active site. The histidine residues, His-132 and His-136, were complexed in their neutral form. It has been reported [20-22] that Ni(II) is complexed by cysteine not through its sulfhydryl hydrogen (- SH), but through the deprotonated negatively charged S - , though it may at first appear unlikely that the –SH proton 17-/14 $58.00+.00 © 2014 Bentham Science Publishers