Studies on HIV-1 core assembly

Abstract

The main objective of this thesis was to define the roles of Gag (p55), in particular, the capsid (CA, p24) protein in human immunodeficiency virus type 1 (HIV-1) particle assembly. More specifically, i) to determine the relative contribution of some specific residues and/or sequences in HIV-1 CA core assembly and virus release, ii) to characterize the importance of two conserved residues with quite opposing intra-molecular contacts with other CA residues in capsid assembly, iii) to define the role and significance of a specific amino acid involved in formation of a conserved β-hairpin structure in HIV-1 capsid assembly, and iv) to define the active antiviral metabolite of an antiviral tripeptide amide previously found to affect HIV-1 capsid assembly and infectivity.

The HIV-1 CA plays a crucial role in both assembly and maturation of the virion. Two highly conserved sequences located in the C-terminal domain (CTD) of HIV-1 CA were investigated with site directed mutagenesis, a valuable technique used widely to study the structure and function of CA in HIV-1 capsid assembly. We showed that mutations of specific residues within the two conserved sequences in the C-terminal domain could affect viral protein expression, virus assembly, release and infectivity. In addition, we showed that these residues are essential for proper proteolytic processing of the Gag/Gag-Pol precursors in a cell-type dependent manner, as well as, for proper morphogenesis of HIV-1 particles.

The importance of two particular amino acid residues, Glu98 and Glu187, located within each of the two CA domains were investigated. In contrast to Glu98 which has no intra-molecular contacts, Glu187 has extensive intra-molecular contacts with eight other CA residues. Furthermore, Glu187 has been shown to be important for a salt-bridge formation in a head-to-tail dimer of HIV-1 CA. We performed detailed analysis to assess the potential effects of mutating these two Glu residues for Ala and Gly, respectively, on Gag processing, virus infectivity, viral cDNA production and virus morphology. In spite of the lack of contact with the other residues of CA as revealed by the structural data, Glu98 was shown herein as a critical element in the action of CA to correctly form mature cores. Our data also showed that the two residues in the study displayed deviated biological properties than the ones being predicted from crystallography and/or analysis of inter-atomic contacts. Thus, Glu187 was found to be dispensable although the residue was predicted to be important for the N-C CA-dimer formation.

As the CA CTD dimer formation is one of the fundamental interactions driving CA multimerization, which also involves hexamerization of the CA N-terminal domains (NTD), we investigated the role and significance of aspartate 51 (D51) which previously has been shown to play a key role in virus assembly and maturation by forming a β-hairpin structure that is highly conserved among retroviruses. In addition to the D51A substitution reported elsewhere, we showed that substitutions of aspartate with glutamate, glutamine, or asparagine, three amino acid residues that are structurally close and have similar properties in protein as aspartate, could not rescue the structural integrity of the protein.

It has previously been shown that addition of the tripeptide GPG-amide, corresponding to a motif found in both conserved sequences described above, could induce non-infectious HIV-1 particles with aberrant core structures. We identified and demonstrated that it is a metabolite of GPG-amide that affects HIV-1 infectivity. This metabolite was purified and its structure determined by NMR to be alpha-hydroxy-glycineamide (alpha-HGA). alpha-HGA binds to the HIV-1 CA and affects its ability to assemble into tubular or core structures in vitro and in vivo. As an antiviral, alpha-HGA has an unusually simple structure, a pronounced antiviral specificity and a novel mechanism of antiviral action.

In conclusion, our findings suggest that mutations in CA are lethal when affecting proper CA core assembly and that the semi-stable non-covalent protein interactions in HIV-1 CA can be specifically disrupted by small molecules, such as alpha-HGA.