FIG 9.
Model of Env conformations on virus particles. (A) Three populations of cleaved Env trimers on virus particles are depicted along with their distinguishing properties. Uncleaved (gp160) Env trimers, which are found to various extents on VLP preparations, are shown on the right. (B) The relative levels of cleaved Env trimer populations on virions are depicted for HIV-1 Envs with different levels of triggerability. The approximate relationship of the different HIV-1 Env variants used in this study is shown. Env triggerability is inversely related to the activation barrier separating State 1 from States 2 and 3 and varies among Envs from different primary HIV-1 strains (107). Envs with intermediate levels of triggerability, like the AD8 Bam Env, populate the pretriggered (State-1) conformation and spontaneously sample more open, downstream (States 2/3) conformations. Viral Envs with lower triggerability (e.g., the Tri Bam or E.1 Bam Envs) populate the pretriggered (State-1) conformation on the virions to a greater extent. Conversely, viral Envs with higher triggerability (e.g., the AD8 Bam 197 HT N Env) exhibit more open trimer conformations on the virions and are more prone to shed gp120, leading to gp41-only trimers. (C) Modulation of virion Env conformational transitions. Upon binding to membrane CD4 on a target cell, the pretriggered (State-1) Env conformation undergoes transitions to more open (State-2/3) intermediates in which the gp120 coreceptor-binding site is exposed. Env binding to the CCR5 or CXCR4 coreceptor promotes additional conformational changes in Env that facilitate virus entry (green arrows). Depending on its triggerability, HIV-1 Env will spontaneously sample more open (State-2/3) conformations (curved black arrow) in which epitopes for some pNAbs become exposed. Such spontaneous transitions from State 1 can be suppressed by State-1-stabilizing Env changes or by treatment of the Env trimers with BMS-806 or chemical crosslinkers (red minus sign). Conversely, State-1-destabilizing Env changes or treatment with sCD4 or CD4-mimetic compounds (CD4mcs) drive Env trimers out of State 1 and increase the level of virion Envs in more open conformations (green plus sign). When near a potential target cell expressing CCR5 or CXCR4 coreceptors, virions with Envs in these open (State-2/3) conformations can infect the cell (168, 169, 171, 186, 187). However, compared with Envs that engage CD4 on a target membrane, Envs opened by other means are more prone to either spontaneous inactivation or neutralization by pNAbs (36, 49, 186, 187, 196). The pNAb-reactive, open Env intermediates exhibit weak intersubunit interactions compared to the bNAb-reactive State-1 Env trimers. Certain HIV-1 isolates like HIV-1AD8 are not neutralized by pNAbs yet show pNAb binding to cleaved Env on virus particles; this apparent paradox can be resolved if the cleaved Env recognized by pNAbs is partially or completely dysfunctional. Our results suggest that gp120 shedding is more likely to occur from more open Env conformers (red arrow).