Improved Protein Mapping for Fragment-Based Drug Design

DESCRIPTION (provided by applicant): Computational solvent mapping methods place molecular probes - small molecules or functional groups - on a protein surface in order to identify the most favorable binding positions. Although X-ray crystallography and NMR show that small organic molecules preferentially cluster in the binding site, current computational methods yield hundreds of energy minima on the surface of the protein, and it is difficult to determine which of the minima are relevant. The Structural Bioinformatics Group at Boston University has developed a novel mapping algorithm that generally eliminates the spurious local minima, and finds the bound positions of small organic probe molecules in good agreement with x-ray or NMR data. The major applications of the method so far have been delineating the active sites of enzymes and other proteins, and detecting minor conformational changes in ligand binding sites. The general goal of the present proposal is to develop this efficient and highly accurate mapping algorithm into the first step of a fragment-based drug design procedure, and apply the method to drug targets that are known to be difficult, thereby developing a scientific and commercial base for collaborative agreements. The Phase I Specific Aims are as follows: (1) modifying the mapping program as required for fragment-based drug design, including improvements in the initial placement of the probes, facilitating the addition of new probes, introducing a more general empirical potential, and improving the evaluation of probe distributions after the mapping; (2) developing optimal fragment libraries by the fragmentation of molecules in databases of pharmaceutically active compounds and then clustering the resulting fragments; and (3) developing a high throughput automated protein mapping software package with appropriate storage, retrieval and analysis of results. We expect that the mapping will correctly place fragments derived from known ligands when mapping either bound or ligand-free protein structures. Using both general and focused fragment libraries, we will explore the binding sites of several important drug targets, including peroxisome proliferator activated receptors (PPPARs), protein tyrosine phosphatase IB (FTP1B), some protein kinases, and cytochrome P450s. Preliminary results suggest that mapping with fragments from well designed libraries will provide very useful information for drug design.