HIV Entry and Its Inhibition Minireview

envelope complexreadily undergoes receptor-activated The human immunodeficiency virus type 1 (HIV-1) is conformational change suggests that its native state is an enveloped virus, and its envelope protein complex metastable, again similar to the pH-activated envelope controls the key process of viral entry. This envelope protein of influenza virus (Carr et al., 1997). That is, protein determines viral tropism and facilitates thememthe labile native state of the HIV envelope complex is brane fusion process that allows invasion of the viral transformed by receptor binding to an energetically genome. The envelope protein can also promote the more stable, fusion-active conformation. fusion of infected cells with uninfected neighboring Protein dissection studies demonstrated that the two cells, a phenomenon called syncytium formation that is 4,3 hydrophobic repeat regions within gp41 form a helireadily observed in cultured cells and may be responsical trimer of antiparallel dimers (Lu et al., 1995). Crystalble for some of the cytopathic effects of advanced HIV lographic analysis confirmed that this gp41 core is a infection. Here we review recent insights in HIV envelope six-helix bundle in which the N and C helices are arprotein structure and function and present our current ranged into three hairpins (Figures 1B and 1C) (Chan et understanding of the entry process. We also review how al., 1997; Weissenhorn et al., 1997). The N peptides form these findings lead to new approaches for inhibiting HIV three central helices arranged in a trimeric coiled coil. entry and may provide insights into the design of better The C peptides form three outer helices that pack in an HIV vaccines. antiparallel manner into highly conserved, hydrophobic The HIV-1 envelope protein complex is initially progrooves on the surface of this coiled coil (Figure 1D). duced as the precursor gp160, which is extensively glyThis structure likely represents the fusion-active conforcosylated and proteolytically cleaved into two subunits mation of gp41 (see discussion in Chan et al., 1997), and by a cellular convertase (Luciw, 1996). The resulting resembles the proposed fusion-active conformations of surface subunit (gp120) and transmembrane subunit the transmembrane envelope proteins from influenza (gp41) remain noncovalently associated and oligomervirus and Moloney murine leukemia virus. ize, most likely as trimers, on the surface of the virion. Fusion Mechanism gp120 binds to CD4 and a coreceptor (a seven-transAny model for HIV entry must account for the remarkable membrane protein of the chemokine receptor family), observation that synthetic C peptides inhibit HIV infecwhich are present on susceptible cells such as T lymtion and syncytia formation at nanomolar concentraphocytes and macrophages. Subsequently, gp41 undertions (Jiang et al., 1993; Wild et al., 1994). Preincubation goes conformational changes that mediate fusion of the of virus with C peptide, followed by its removal, does viral membrane with the target cell membrane. HIV entry not block HIV entry; this feature suggests that these is therefore a complex process involving multiple protein peptides do not act on the native conformation of gp41. interactions, each of which is a potential target for the Based on the structural features of the fusion-active development of antiviral compounds. gp41 core, these peptide inhibitors likely work by bindgp41 Structure ing to the trimeric coiled coil of gp41, thereby acting The gp41 molecule is a transmembrane protein with through a dominant-negative mechanism (Lu et al., several important features within its ectodomain (Figure 1995; Chan et al., 1997; Weissenhorn et al., 1997). 1A). First, the amino terminus of gp41, created by proA simple dominant-negative model, however, fails to teolytic cleavage of the gp160 precursor, contains a explain adequately one puzzling feature of C peptide hydrophobic, glycine-rich “fusion” peptide that is esseninhibition: How can C peptides act at such low concential for membrane fusion. Second, there are two regions trations, given that the N and C peptide regions are with a 4,3 hydrophobic (heptad) repeat, a sequence mowithin a single gp41 molecule? That is, how can the C tif characteristic of coiled coils. Synthetic peptides (see peptides so effectively inhibit an intramolecular associabelow) derived from these two regions are termed N tion reaction, in which the two regions of gp41 are pres(amino-terminal) and C (carboxy-terminal) peptides. Beent in a high effective concentration? It seems likely that tween these two heptad repeat regions is a loop region the solution to this puzzle is that C peptides must bind containing two cysteines. to gp41 prior to formation of the six-helix complex beA large number of studies support the notion that the cause, once this gp41 core is assembled, it is extremely envelope complex exists in at least two major conformastable (the melting temperature of the gp41 core is in tions (see references in Chan et al., 1997). The native, excess of 908C) and is unlikely to be disrupted by exogeor nonfusogenic, conformation exists on the surface