Characterisation and Inhibition of the Retroviral HIV-Protease

Abstract

Virus encoded proteases were first described in avian and murine retroviruses (von der Helm, 1977; Yoshinaka and Luftig, 1977). They are ‘processing’ proteases cleaving virus encoded polyproteins into smaller proteins which then function as viral structural (scaffolding) proteins or virus specific enzymes. Thus, they are anabolic rather than catabolic enzymes. They have been demonstrated in many viruses different from retroviruses (for review see Kräusslich and Wimmer, 1988). In retroviruses they were shown to process the gag-pol protein precursor to yield proteins required for production of infectious virus. It has been shown that replication of murine retrovirus with a defective protease yielded immature virus particles lacking infectivity (Crawford and Goff, 1985; Katoh et al., 1985). Replication of HIV (human immunodeficiency virus) carrying a lethal protease mutant led to non-infectious virus particles (Kohl, et al., 1988). Retroviral proteases appear to have a highly specific relation to their virus encoded substrate (Dittmar and Moelling, 1978; Khan and Stephenson, 1979; Yoshinaka et al., 1985; Kräusslich and von der Helm, 1987). Examination of the amino acid sequences of retroviral proteases indicated the presence of a conserved Asp-Thr(Ser)-Gly sequence, suggesting that they are aspartic-type proteases (Kräusslich and Wimmer, 1988; Toh et al., 1985; Pearl and Taylor, 1987). By site specific mutagenesis of the essential Asp-25 of the HIV protease (PR) performed by recombinant technique we and others could prove directly that the PR has an aspartic proteolytic site (Kohl et al., 1988; Mous et al., 1988; Seelmeir et al., 1988; Loeb et al., 1989) because the Asp-25 mutation rendered the enzyme inactive. The enzyme is proteolytically active only as a (homo)-dimer of two of the genetically encoded monomers (Meek et al., 1989; Katoh et al., 1989) as theoretically predicted (Pearl and Taylor, 1987).