Ligand-binding sites on the Mycobacterium tuberculosis urease

Abstract

Introduction.Mycobacterium tuberculosis is the causative agent of tuberculosis that remains a serious medical and social health problem. Despite intensive efforts have been made in the past decade, there are no new efficient anti-tuberculosis drugs today, and that need is growing due to the spread of drug-resistant strains of M.tuberculosis.M. tuberculosis urease (MTU), being an important factor of the bacterium viability and virulence, is an attractive target for anti-tuberculosis drugs acting by inhibition of urease activity.However, the commercially availableurease inhibitors are toxic and unstable,that prevent their clinical use. Therefore, new more potent anti-tuberculosis drugs inhibiting new targets are urgently needed. A useful tool for the search of novel inhibitors is a computational drug design. The inhibitor design is significantly easier if binding sites on the enzyme are identified in advance. This paper aimed to determine the probable ligand binding sites on the surface of M. tuberculosis urease. Methods. To identify ligand binding sites on MTU surface, сomputational solvent mapping methodFTSite was applied by the use of MTUhomology model we have builtearlier. The method places molecular probes (small organic molecules containing various functional groups) on a dense grid defined around the enzyme, and for each probe finds favorable positions. The selected poses are refined by free energy minimization, the low energy conformations are clustered, and the clusters are ranked on the basis of the average free energy. FTSite server outputs the protein residues delineating a binding sites and the probe molecules representing each cluster. To predict allosteric pockets on MTU, AlloPred and AlloSite servers were applied. AlloPred uses thenormal mode analysis (NMA) and models how the dynamics of a protein would be altered in the presence of a modulator at a specific pocket.Pockets on the enzymeare predicted using the Fpocket algorithm. To model the reduction in flexibility of allosteric pocket on modulator binding, the unperturbed normal modes are first calculated for the protein. The calculation is then repeated, each time perturbing one of the pockets in the protein.These results are combined with output from Fpocket in a support vector machine (SVM) to predict allosteric pockets on proteins.The AlloSite serveris similar to the AlloPred method in that it uses the Fpocket algorithm to elucidate allosteric pockets, whereas AlloPred uses an approach that combines flexibility with the Fpocket output. Results and discussion.By computational solvent mapping method FTSite,we have exploredM.tuberculosis urease nonamer surface tofind sites that tend to bind small organic molecular probes representing fragments of drug molecules with diverse hydrophobic and hydrophilic properties. The predicted three top ranked binding sites were situated at the interfaces between chains C and A, and chain G of neighbour trimer (and at equivalent locations in symmetrical trimers as well). A mapping of enzymes generally yields the most probable sites situated in a subsite of the enzyme active site. This was not the case for MTU which active sites were inaccessible for probes due to the closed conformation of the covering flap, and predicted binding sites were located not far from them at the entrance into a deep pocket. To explore their possible structural and functional role, we correlated the locations of predicted MTU binding sites and its ancillary pockets (which remain open andsolvent exposed even while the flap is closed) and indicated their partial overlapping. This overlapping may suggest that predicted sites are likely the intermediate binding sites responsible for recruiting a ligand to another binding site deeply buried in the protein. To examine the possibility that predicted binding sites are the sites for allostery binding we carried out the search for probable sites of allostery binding on MTU surface by AlloPred and AlloSite servers. Predicted probable allosteric sites overlapped with binding sites revealed by FTSite suggesting their possible function as sites for allosteric binding. Conclusions. On the surface of M.tuberculosis urease, there were revealed theprobable ligand binding sites that appear to be the sites of allosteric binding. They may serve as promising targets for designing novel allosteric modulators as receptor-selective anti-tuberculosis drugs.