Structure of the aspartic protease from Rous sarcoma retrovirus refined at 2-.ANG. resolution

Publisher Website
Google Scholar
Abstract
The structure of Rous sarcoma virus protease has been solved by multiple isomorphous replacement in the crystal form belonging to space group P3121, with unit-cell parameters a = 88.95 .ANG. and c = 78.90 .ANG.. The enzyme belongs to the family of aspartic proteases with two identical subunits composing the active homodimer. The noncrystallographic dyad relating these two subunits was identified after preliminary tracing in the MIR map and was used for phase improvement by electron-density averaging. Structure refinement resulted in a model that included 1772 protein atoms and 252 water molecules, with an R factor of 0.144 for data extending to 2 .ANG.. The secondary structure of a retroviral protease molecule closely resembles that of a single domain in pepsin-like aspartic proteases and consists of several .beta.-strands and of one well-defined and one distorted .alpha.-helix. The dimer interface is composed of the N- and C-terminal chains from both subunits which are intertwined to form a well-ordered four-stranded antiparallel .beta.-sheet. In each monomer, the catalytic triad (Asp-Ser-Gly) is located in a loop that forms a part of the .psi.-structure characteristic to all aspartic proteases. The position of a water molecule between the active-site aspartate residues and the general scheme of H bonding within the active site bear close resemblance to those in pepsin-like aspartic proteases and therefore suggest a similar enzymatic mechanism. The binding cleft over the active site is covered by two flap arms, one from each monomer, which are partially disordered. The retroviral protease dimer has been compared with several enzymes of cellular origin, with chains aligning to an rms deviation of 1.90 .ANG. or better.