Dr. Furey’s research involves the structure determination and analysis of large biological molecules and complexes by x-ray crystallography, and correlating the results with known functions. The work currently focuses on thiamin (vitamin B1) dependent enzymes and cell cycle regulating enzymes, as well as crystallographic methods development. Results of these studies could lead to development of therapeutic agents directed against pathogenic organisms, and anti-cancer drugs.
The pyruvate dehydrogenase multienzyme complex (PDHc, MW 4.7 million Daltons, 60 protein subunits & 60 active sites for the E. coli version) is present in most organisms and is critical for carbohydrate metabolism where it converts pyruvate, the product of glycolysis, to acetyl-CoA via a complicated process of substrate channeling within the confines of the complex. Structural analyses of the complex and its three major enzymatic components E1 (24 copies), E2 (24 copies), & E3 (12 copies) are underway, and Dr. Furey has already determined high resolution crystal structures for some of the components and reaction intermediates from the E. coli version.
The E1 components are rate determining and require thiamin diphosphate as a cofactor, but must interact with a flexible segment [lipoyl domain (LD) and associated lipoamide side chain] on an E2 to transfer the first reaction product, an acetyl group, to the E2 active site. The acetyl group is then transferred to co-enzyme A within the E2 active site, and the product acetyl-CoA is released. The E2 bound lipoamide group (now reduced) then moves to an E3 (FAD dependent) active site, where it is oxidized to restore the initial conditions. Binding of the flexible segment to E1 and E3 subunits is mediated by additional binding to a peripheral subunit-binding domain (PSBD), shown bound to E1.
Mechanistic details regarding the catalytic reactions in each active site are sought, as well as identifying structural aspects critical for assembly of the individual components to form the complete multienzyme complex. Specific mutations in some of the components are associated with hereditary diseases in humans, and detailed analyses of the structure-function relationships may suggest development of plausible therapeutic agents to counter the effects of the mutations. Additionally, given the critical nature of this system in overall energy production for cellular function, development of inhibitors binding at any of the catalytic sites, or at sites disrupting protein-protein assembly, may considerably weaken or kill the organism. Lack of appreciable sequence homology between PDHc’s from humans and pathogenic bacteria therefore suggests that effective, pathogen specific antibacterial agents may be developed.
Early expression or over expression of Cdc25 proteins can cause the cell to prematurely progress leading to oncogenic effects, making these enzymes exciting targets for anti-cancer drug development. As part of a collaborative effort with Dr. John Lazo’s group, several potent inhibitors of Cdc25 proteins have been discovered, and structural analysis of their complexes with the enzymes are underway to reveal both where and how these inhibitors function. Dr. Furey has crystallized the catalytic domain of Cdc25b and determined its high-resolution structure. His group is currently co-crystallizing the catalytic domain with several inhibitors, as a step towards development of effective anti-cancer agents via structure-based drug design procedures.
In collaboration with the Hauptman-Woodward Institute for Medical Research, Dr. Furey's group is developing new computational methods for solving macromolecular crystal structures by automated techniques. This work involves creating and developing a software package BnP, which is a merging of the PHASES package developed in the Furey lab, and the SnB package developed in Buffalo. A simple, graphical user interface is developed to enable automatic creation of an interpretable electron density map starting from observed x-ray diffraction data, with only a few mouse clicks and text field entries required. This will invoke automatic scaling of data, determination of heavy atom/anomalous scatterer sites, refinement and validation of sites, calculation of protein phases, phase refinement, and phase improvement via solvent flattening/negative density truncation. A few more mouse clicks enable automated building of a complete or nearly complete model by interfacing with other externally developed software. The idea is to make it simple for novices to determine good quality crystal structures, while enhancing the productivity of more sophisticated users as well.