Research Focus: Assembly and Release of HIV-1 and Other Retroviruses
Retroviral Gag proteins are synthesized in the cytoplasm of the infected cell and assemble into virus particles that typically bud from the plasma membrane (PM). Expression of Gag proteins alone is generally sufficient for the assembly and release of noninfectious, virus-like particles (VLPs). The mature Gag proteins [matrix (MA), capsid (CA), and nucleocapsid (NC)] are generated concomitant with virus release upon cleavage of the Gag precursor by the viral protease (PR) [Freed, Nat. Rev. Microbiol. 13: 484-496, 2015]. PR-mediated Gag processing leads to virus maturation, a morphological transition essential for virus infectivity. Retroviral gag genes often encode other domains and spacer peptides in addition to MA, CA, and NC. For example, the HIV-1 Gag precursor includes two spacer peptides (SP1 and SP2) and the p6 domain.
Retroviral Gag trafficking and assembly. After Gag synthesis, the MA domain directs Pr55Gag to the PM. The affinity of the MA domain for membrane is provided in part by a myristic acid moiety covalently attached to the N-terminus of MA. Sequences in MA downstream of the myristate also contribute to membrane binding—in particular, a highly basic patch of amino acid residues. Structural analysis of myristylated MA has indicated that the myristate moiety adopts both an exposed and a sequestered conformation and that the degree of myristate exposure regulates Gag–membrane association.
A large body of data indicates that HIV-1 assembly takes place predominantly at the PM [Freed, Nat. Rev. Microbiol. 13: 484-496, 2015]. In monocyte-derived macrophages (MDMs), assembly and budding can occur in an internal compartment that apparently represents deep, specialized invaginations that are connected to the PM via nanoscale channels [Bennett et al., PLoS Pathog. 5: e1000591, 2009]. We reported that single amino acid changes in the highly basic domain of MA cause virus assembly to be redirected to late endosomes or multivesicular bodies (MVBs). A focus of our work has involved characterizing the mechanism of virus spread from infected MDMs to T cells [Gousset et al., PLoS Pathog. 4: e1000015, 2008].
Although the cellular determinants that regulate the site of HIV-1 assembly remain to be fully defined, we demonstrated that lipid rafts serve as sites for assembly at the PM [Ono & Freed, PNAS 98: 13925-13930, 2001] and identified the phosphoinositide lipid phosphatidylinositol-(4,5)-bisphosphate [PI(4,5)P2] as a host factor involved in directing Gag to the PM. Depleting PI(4,5)P2 leads to the retargeting of HIV-1 assembly to MVBs and severely disrupts virus particle production [Ono et al., PNAS 101: 14889-14894, 2004]. Structural studies have shown that HIV-1 MA, as well as the MA domain of several other retroviruses, interacts directly with PI(4,5)P2 and that some of the basic residues we identified as being important for Gag targeting to the PM engage in electrostatic interactions with PI(4,5)P2. We hypothesize that host factors in addition to PI(4,5)P2 play a vital role in the trafficking of Gag to the PM. For example, we have demonstrated that the ADP ribosylation factors (Arfs) and Golgi-localized, gamma-ear-containing, Arf-binding (GGA) proteins function in directing Gag to the PM and modulating virus budding [Joshi et al., Mol. Cell 30: 227-238, 2008], as does the cellular SNARE pathway [Joshi et al., J. Biol. Chem. 286: 29861-29871, 2011]. We are currently working to define the role of host cell machinery in Gag trafficking to the PM and will seek to develop inhibitors of HIV-1 Gag trafficking, membrane binding, and assembly. In this project, we will also define viral and cellular determinants involved in directing HIV-1 assembly to the virological synapse, from which cell–cell transfer efficiently takes place.
Env incorporation. The envelope (Env) glycoproteins of HIV-1 are synthesized as a polyprotein precursor, gp160, that is proteolytically processed by a cellular protease to generate the surface (SU) glycoprotein gp120 and the transmembrane (TM) glycoprotein gp41. Incorporation of Env glycoproteins into virions is an essential step in the replication process; however, the mechanism by which the Env glycoproteins are incorporated remains incompletely characterized. Several lines of evidence suggest that HIV 1 Env glycoproteins are actively recruited into virions via direct interactions between Env and MA; for example, mutations in both the MA domain of Gag and the cytoplasmic tail of gp41 can block HIV 1 Env incorporation. We showed that in most cell lines and in primary cell types, gp41 truncations severely disrupt Env incorporation but that in some cell lines these gp41 truncation mutants are efficiently incorporated [Murakami & Freed, PNAS 97: 343-348, 2000]. The cell-type-dependent requirement for the gp41 cytoplasmic tail in Env incorporation hints at the involvement of host cell factors. However, the identity of such factors remains to be defined. We are currently using a range of cell biology, virology, biochemical, and imaging approaches to characterize viral and cellular determinants of Env incorporation. Our recent findings highlight a requirement for MA trimers in Env incorporation [Tedbury et al., PNAS 113: E182-E190, 2016]. We are also defining host cell factors that play a role in HIV-1 Env glycoprotein incorporation [Checkley et al., J. Virol. 87: 3561-3570, 2013].
Retrovirus budding. The p6 domain of HIV-1 Gag is required for efficient virus budding. We mapped the virus release function of p6 to a highly conserved Pro-Thr-Ala-Pro (PTAP) motif near the N-terminus of p6 [Fujii et al., Nat. Rev. Microbiol. 5: 912-916, 2007]. Other retroviral Gag proteins also possess “late domains” that, like the PTAP motif of p6, promote virus release. The HIV-1 accessory protein Vpu also promotes particle release after the completion of budding by counteracting the host cell restriction factor tetherin.
Numerous lines of evidence support the hypothesis that retroviral late domains function by interacting with host factors—specifically, components of the cellular endosomal sorting complex required for transport (ESCRT) machinery [Freed, Nat. Rev. Microbiol. 13: 484-496, 2015]. The ESCRT machinery is composed of four multiprotein complexes (ESCRT-0, I, II, and III) and a variety of accessory proteins including Alix and the ATPase Vps4. This cellular apparatus functions in the sorting and release of cargo proteins into vesicles that bud into late endosomes and also plays a key role in the abscission step of cytokinesis. Retroviruses have evolved to usurp this cellular budding machinery to promote their release from the PM. Work from several labs, including ours, demonstrated that HIV-1 recruits the ESCRT machinery primarily through a direct interaction between the PTAP motif in p6 and the ESCRT-I component tumor susceptibility gene 101 (Tsg101). Secondary interactions occur between p6 and Alix. We are seeking to elucidate a number of key issues relating to lentivirus budding and are attempting to develop small-molecule inhibitors of the p6–Tsg101 interaction to disrupt HIV-1 budding. Of particular interest is the identification of cellular partners for the ESCRT-associated protein Alix [Keren-Kaplan et al., EMBO J. 32: 538-551, 2013].
Virion maturation. PR-mediated cleavage of the Gag and Gag–Pol precursors leads to a dramatic change in virion morphology, a process known as maturation. The highly ordered nature of Gag processing and the strict dependence on complete processing for proper virion maturation make the Gag processing cascade a potential target for drug development. Indeed, the betulinic acid derivative dimethylsuccinyl betulinic acid [PA-457 or bevirimat (BVM)] potently inhibits HIV-1 infectivity by targeting a late Gag processing event: the cleavage of the CA–SP1 processing intermediate to mature CA [Li et al., PNAS 100: 13555-13560, 2003]. By specifically disrupting this step in Gag processing, BVM treatment leads to the formation of poorly infectious viral particles with aberrantly condensed cores. We have selected and characterized a number of single amino acid mutations in the CA–SP1 boundary region that confer resistance to BVM, establishing this region of Gag as the target of the inhibitor [Adamson et al., J. Virol. 80: 10957-10971, 2006]. Clinical trials conducted with BVM provided mixed results; in some patients significant reductions in viral loads were achieved, whereas in other patients no benefit of BVM therapy was observed. We and others were able to attribute this lack of response to polymorphisms in SP1. We have recently initiated studies on a second HIV-1 maturation inhibitor, PF-46396, developed by Pfizer. While structurally distinct from BVM, PF-46396 also inhibits CA–SP1 processing. Ongoing studies will define the target and mechanism of action of PF-46396. Particularly interesting is our observation that a number of PF-46396-resistant mutants are severely defective for assembly and infectivity in the absence of the compound but produce infectious virions in its presence [Waki et al., PLoS Pathog. 8(11): e1002997, 2012]. We are also engaged in studies aimed at elucidating the structure of the maturation inhibitor-binding site in assembled Gag. We believe that increased structural information on the HIV-1 maturation inhibitor-binding site will lead to the discovery of new, structurally distinct inhibitors. Finally, with collaborators at DFH Pharma, we have identified a series of highly potent BVM derivatives that display broad activity against strains of HIV-1 that are resistant to BVM [Urano et al., Antimicrob. Agents Chemother. 60: 190-197, 2016]. This work advances our efforts to develop virus maturation inhibitors as an antiviral strategy.
modified: 11 April 2017