Immature HIV-1 assembles from Gag dimers leaving partial hexamers at lattice edges as potential substrates for proteolytic maturation

The CA (capsid) domain of immature HIV-1 Gag and the adjacent spacer peptide 1 (SP1) play a key role in viral assembly by forming a lattice of CA hexamers, which adapts to viral envelope curvature by incorporating small lattice defects and a large gap at the site of budding. This lattice is stabilized by intrahexameric and interhexameric CA-CA interactions, which are important in regulating viral assembly and maturation. We applied subtomogram averaging and classification to determine the oligomerization state of CA at lattice edges and found that CA forms partial hexamers. These structures reveal the network of interactions formed by CA-SP1 at the lattice edge. We also performed atomistic molecular dynamics simulations of CA-CA interactions stabilizing the immature lattice and partial CA-SP1 helical bundles. Free energy calculations reveal increased propensity for helix-to-coil transitions in partial hexamers compared to complete six-helix bundles. Taken together, these results suggest that the CA dimer is the basic unit of lattice assembly, partial hexamers exist at lattice edges, these are in a helix-coil dynamic equilibrium, and partial helical bundles are more likely to unfold, representing potential sites for HIV-1 maturation initiation.

The polyprotein Gag is the main structural component of HIV-1, consisting of the MA (matrix), CA (capsid), NC (nucleocapsid), and p6 domains as well as the spacer peptides SP1 and SP2 (1). Gag is produced in infected host cells and trafficked to the plasma membrane, where it assembles into a hexagonal lattice via its CA domain and recruits other viral proteins and the viral RNA genome (1, 2). Assembly of the curved Gag lattice is commensurate with membrane bending at the site of assembly, after which recruitment of Endosomal Sorting Complex Required for Transport III (ESCRT-III) components by the p6 domain of Gag induces membrane scission and release of the immature virus particle (2). The hexagonal Gag lattice accommodates curvature in the growing bud by incorporating vacancy defects (3). The activity of ESCRT-III is timed such that the final immature lattice is incomplete, giving rise to an additional large gap in the lattice, resulting in a truncated spherical shape (4–6).During or after budding, the viral protease is activated and cleaves this immature Gag lattice into its component domains, which leads to structural rearrangement within the virus particle (2). The released CA domains assemble to form a closed, conical capsid around the condensed ribonucleoprotein (RNP) complex of the mature virus (1, 7). Maturation is required for the virion to become infectious (1).Within the immature virus particle, the N-terminal domain of CA (CA_NTD) forms trimeric interactions linking three Gag hexamers while the C-terminal domain of CA (CA_CTD) forms dimeric interactions mediated by helix 9 of CA, linking two Gag hexamers together (8). The CA_CTD additionally forms intrahexamer interactions around the sixfold axis of the hexamer (8, 9). Amphipathic helices formed by the C-terminal residues of CA_CTD and the N-terminal residues of SP1 junction assemble into a six-helix bundle (6HB), thereby imposing hexagonal order on the CA domains, via classical knobs-in-holes packing mediated by exposed hydrophobic side chains, as also seen in coiled coils (9, 10). In combination, these relatively weak interactions give rise to a very dynamic, reversible assembly process that prevents the assembling lattice from becoming trapped in kinetically unfavorable states (11). This robust assembly behavior is consistent with icosahedral viruses (12–15). It is not surprising, therefore, that the energetics of Gag assembly are tightly controlled and highly dependent on scaffolding effects from the viral RNA and the membrane-interacting MA domain of Gag in order to ensure productive viral assembly (11, 16). Analysis of the diffusion pattern of fluorescently labeled Gag supports the notion that Gag is trafficked to the site of assembly as low-order multimers, although it is still unclear whether these are Gag dimers, trimers, or other multimeric forms of Gag (16, 17).The primary assembly unit of the Gag lattice remains largely unknown. We can identify two hypothetical ways in which the lattice could assemble. First, the lattice could grow by addition of Gag hexamers (or sets of six component monomers), such that the CA-SP1 junction is assembled within a hexameric 6HB at all positions in the lattice. In this case, interfaces between hexamers would be unoccupied at the edge of the lattice. From a purely energetic perspective, this appears most reasonable. Second, the lattice could form via addition of Gag dimers or Gag trimers (or equivalently from sets of either two or three component monomers). This would maintain, for example, the dimeric CA-CA interhexamer interactions but leave incomplete hexamers at the lattice edges, including unoccupied hexamer-forming interfaces along the CA-SP1 bundle. It additionally remains unclear whether the unoccupied Gag-Gag interfaces at the lattice edges are simply exposed, or whether they are stabilized by alternative conformations of individual domains or proteins, or by other binding partners. Understanding the structure of the edge of the immature Gag lattice therefore has implications for understanding the mechanism of virus assembly.Viral assembly, budding, and maturation are tightly linked, and disrupting the kinetics of any of these processes can give rise to defects in maturation and formation of noninfectious viral particles (1, 18, 19). The rate-limiting proteolytic cleavage site in the maturation process resides within the CA-SP1 6HB (20). Unfolding of the helical bundle is required to allow proteolytic cleavage to proceed (21–23), but the exact mechanism for protease access to this site is not known. The spatial localization of proteolytic processing within the context of the immature Gag lattice is relevant: Does the protease act on Gag within the lattice, or does it act on the edges of the Gag lattice, causing a cascade of lattice disruption? At the lattice edge, is the substrate for the protease with a 6HB or within an incomplete hexamer? Understanding the structure of the edge of the immature Gag lattice therefore has implications for understanding the mechanism of virus maturation.High-resolution immature Gag structures have previously been determined directly from purified viruses by cryo-electron tomography (cryo-ET) and subtomogram averaging (10). These structures represent an average hexamer within the immature lattice, with a full complement of six Gag hexamer neighbors. Here, we have applied subtomogram classification and averaging approaches to an existing immature virus dataset (10) in order to determine the structures of Gag assemblies at lattice edges. We also applied atomistic molecular dynamics simulations to assess the roles of the different CA-CA interactions in immature lattice stabilization and predict the properties of the structures we observe at lattice edges. Together, our results suggest that the basic unit of immature HIV-1 assembly is a Gag dimer and partial CA-SP1 helical bundles are present at the edges of the assembled lattice and may be substrates for initiation of maturation.