A Novel Conformation of the ZIKV Protease Provides a Unique 3D Scaffold for Discovering Allosteric Protease Inhibitors as Direct Anti-virals, via HT and Virtual Screening, and Protein Engineering

点击翻译按钮获取中文摘要

Project Summary Zika flavivirus (ZIKV) is an emerging pathogen globally and in the US. ZIKV infection usually causes mild symptoms; however, ZIKVBR infection has in some individuals resulted in serious neurological sequelae, including Guillain-Barré syndrome in adults, and microcephaly in prenatally-infected infants. No treatments for flaviviral infections are available. Vaccine strategies for DENV flavivirus have been disappointing, possibly due to Antibody-Dependent Enhancement of infection (ADE); and it has been demonstrated that antibodies to DENV result in ADE for ZIKV. Furthermore, ZIKV is present at very high levels in semen for up to six months post-infection. Since only ~20% of infected persons exhibit visible signs of infection, the risk of sexual transmission is greatly increased. These issues highlight the need for therapeutic small-molecule approaches that directly target the viral life-cycle. The flaviviral NS2B-NS3 protease is required for polyprotein cleavage and viral infectivity and represents an attractive target. The major innovation of this proposal arises from our recent crystallographic observation of a novel “third” conformation of ZIKV NS2B-NS3 protease, which we call the “double-open” state. It is distinguished from the “closed” and “single-open” states by a radical reorganization of the C-terminal substrate-binding β-hairpin of NS3PRO that is incompatible with protease activity. This double- open state displays a new surface-exposed, deep hydrophobic pocket, distal to the active site, which appears highly druggable, since it is highly conserved among flaviviruses, and lined with elements of the reorganized β- hairpin. Thus, small molecules that bind tightly to this pocket should stabilize this inactive state, and act as allosteric inhibitors. We propose to test this hypothesis starting with both real and virtual libraries. In principle, allosteric inhibitors should have much greater specificity compared with active-site inhibitors, since many host proteases have very similar, or identical target recognition sequences. We further hypothesize that our unique scaffold should provide a superior pathway to refinement of inhibitory hits by generating co-crystal structures to guide further design, an element that has been problematic for the development of more conventional allosteric inhibitors that target shallow pockets at the labile NS2B-NS3 interface. The nature of our novel pocket should expedite co-crystallization. We will screen compound libraries using both conventional HT approaches and pocket-directed in silico screening. The primary screen is PTS, with mutant protease locked into the “double- open” state. Small-molecule pocket binders should raise the melting temperature. Host proteases with similar target preferences will be used as counter-screens. Protease inhibition by surviving hits will be tested using a HT fluorescent peptide screen. HepaRG cells will be used to test for cytotoxicity and Huh-7.5 cells for inhibition of infection by ZIKV Brazil strain. We will validate allosteric binding using biophysical techniques, co- crystallization and enzymology; and direct inhibitor optimization by co-crystallization. We anticipate obtaining ~5 inhibitors suitable for further therapeutic development.