Assembly 

The Gag protein is central to HIV assembly.  It recruits the polymerase and the RNA into the capsid structure and participates in the association of the nucleocapsid with the cytoplasmic domain of the gp41 protein on the inner surface of the plasma membrane. 

The primary translation product of the GAG gene is a polyprotein of 55kD (Gag) which is eventually cleaved when the virus matures during of after budding. 55kD Gag associates with the inner surface of the plasma membrane at its N-terminus (the MA domain) which is linked covalently to a fatty acid (myristic acid) that embeds in the lipid bilayer (figure). The MA domain also contains basic amino acids that interact (non-covalently) with negatively charged phosphate groups of membrane phospholipids. A second functional domain of Gag protein is located at the C-terminus of what will become the p24 capsid protein and is important is assembly of the virus and its budding from the cells.  A third important part of Gag is located at the C-terminus of the capsid (p24) domain and the N-terminus of the nucleocapsid (p7) domain. This region of the protein also contains numerous basic amino acids and associates with other Gag proteins (to form multimeric complexes) and with the polymerase. There are two zinc binding sequences of amino acids here and they, together with the basic residues, associate with the viral genomic RNA. 

In most cases, translation of the full size genomic RNA terminates at the end of the GAG gene but in about one occasion in twenty, as a result of a frame shift, translation continues into the POL gene to form a Gag-Pol polyprotein.  

Assembly takes place is a sequence of stages:

-         The uncleaved (p55) Gag polyprotein forms multimeric complexes involving the C-terminus of the CA domain and the N-terminus of the NC domain. This is crucial to the following steps

-         Gag polyprotein multimers bind to two molecules of the genomic RNA

-         The Gag-Gag multimers bind to Gag-Pol polyproteins in a ration of about twenty to one.

-         Vpr associates with the Gag-Gag multimers. It binds to the p6 (C-terminal) domain of Gag. There is about one Vpr molecule for every seven Gag polyproteins.

-         Vif associates with the complex of Gag-Gag, Gag-Pol and RNA. If the RNA does not bind, neither does Vif. Vif has affinity for the RNA but also for Gag

-         Complexes associate with the inner surface of the plasma membrane, possibly in discrete regions known as lipid rafts. This involves the myristic acid and the membrane associating region at the N-terminus MA domain of both the Gag-Gag polyprotein monomers and the Gag-Pol polyprotein. Lipid rafts are detergent-insoluble regions of the plasma membrane in which there is a concentration of sphingomyelin, other glycosphingolipids and cholesterol together with glypiated proteins. Since this is the region from which the virus buds, it is not surprising that the lipids of the HIV-1 envelope reflect the composition of the raft rather than the plasma membrane as a whole.  

Several cell proteins also associate with the nucleocapsids and three have been investigated. Cyclophilin A binds to Gag and will be involved in the infectivity of the virus when it reaches a new host cell. HP68, which is an RNase inhibitor, associates with Gag, Gag-Pol and Vif and is involved in nucleocapsid assembly. A third protein called tumor susceptibility gene 101 (Tsg101) interacts with the C-terminal p6 domain and is involved in budding. It has some function in the sorting of endosomes (and therefore in vesicle budding from membranes) and in ubiquitin-binding, a property that also seems to be important in the budding of the virus; indeed, Tsg101 may bind to Gag as a result of Gag ubiquitinylation 

Budding

The complexes of Gag, Gag-Pol and RNA, together with lesser amounts of the smaller viral proteins and host cell proteins, associate with the lipid rafts at the cell surface. These can be clearly seen as darkly stained areas in electron micrographs. The binding of the N-terminal regions of Gag to the membrane lipids induces curvature of the bilayer. Tsg101 is also involved in this.  

Maturation

The electron-dense layer containing the Gag complexes that first appears on the inner surface of the plasma membrane is found in the budding virions as a doughnut-shape in which the Gag proteins are arranged in a radial pattern with the N-terminal myristic acid pointing outwards. The appearance changes as the virus matures to form the cone-shaped nucleocapsid structure (figure). The transition from doughnut to cone results from the proteolytic cleavage of Gag by the Pol domain of the Gag-Pol polyprotein which is necessary for the new virus particle to infect another cell.  First, the Gag-Pol protein undergoes auto-proteolysis, liberating the protease. The protease is a dimer and comes from a dimer of Gag-Pol.  Next, the Gag protein is cleaved between p2 and NC; second, between MA and CA; and finally between CA and p2. Cleavage is necessary for the condensation of the nucleocapsid; that is packing of the RNA into the NCp7 core of the virus, and the formation of an outer shell containing CA p24.