Targeting the HIV-1 RNA packaging signal via small molecules

Thomas L. James, University of California, San Francisco
Molecular Biology

Despite the success of highly active antiretroviral therapy (HAART), great concern remains due to rapid emergence of new HIV-1 genetic variants that help the virus escape host immunity and multidrug therapies. We propose a project with the long-term goal of finding a new anti-HIV drug that is less susceptible to evasion by resistant HIV-1 strains. Specifically, we will explore the feasibility of attacking a novel target in the HIV-1 RNA, which is nearly 100% conserved. That target is in the HIV-1 RNA packaging signal, and the aim is to inhibit packaging of RNA into a new virion using specially designed small molecules. The stem-loop 1 SL1 of the packaging RNA signal carries a G-rich internal loop (GRIL), whose sequence is nearly 100% conserved. The GRIL motif binds the nucleocapsid protein (NCp7). Mutations in this GRIL impair packaging of RNA and overall infectivity of the virus. Recently, we solved a high-resolution NMR structure of the entire SL1 RNA including the very highly conserved GRIL. We intend to initiate structure- based discovery of ligands that will bind to the GRIL RNA and inhibit NCp7 interacting with this site. Any ligands discovered will be selected for good bioavailability and no obvious toxicity, with promise for development into good drug leads. The rate of mutations in GRIL is apparently very low, so emergence of HIV-1 strains resistant to any drug that binds will be very slow. For this two-year project, we will obtain ligands that bind well to the GRIL motif, which will subsequently be tested and optimized for reducing RNA packaging and viral infectivity. To achieve the goals of this two-year project, we will computationally screen a database of "drug-like" molecules, and then experimentally test 600 hits for their binding to the GRIL RNA and for inhibition of NCp7 binding. For computational screening, we will use novel methodology: docking with our recently developed program MORDOR. MORDOR has many new docking features, the most important for this project being a very efficient sampling algorithm and the ability to achieve an induced fit of ligand to receptor permitting flexibility in target RNA conformation during docking. Allowing target flexibility is crucial for RNA targets, because unpaired residues in RNA (e.g., bulges or loops) are usually flexible and often adapt to bind a ligand. Compounds emerging from the virtual screening will be tested in vitro for their binding to the target RNA using two NMR techniques. One is saturation transfer difference, which we have adapted specifically for RNA targets. It is relatively quick in telling if a ligand binds to an RNA target and in mapping of the binding epitope on the ligand. The second technique entails monitoring RNA proton chemical shifts upon addition of increasing concentrations of a ligand. This method identifies a specific epitope on the RNA where a ligand binds, and also distinguishes specific from non-specific binders. NMR chemical shift titrations also enable determination of the binding affinities of ligands. Compounds that bind specifically to the GRIL motif of SL1 RNA will be tested in vitro for their inhibition of NCp7 protein binding to the same target. For this purpose, we will record quenching of the fluorescence of the sole tryptophan of NCp7 by its binding to RNA in the presence of an increasing concentration of a potential inhibitor. Finally, the best compounds will be tested for their inhibition of viral growth.