The HIV-1-envelope (Env) spike, comprising three gp120 and three gp41 subunits, is a conformational machine that facilitates HIV-1 entry by rearranging from a mature unliganded state, through receptor-bound intermediates, to a postfusion state. As the sole viral antigen on the HIV-1-virion surface, Env is both the target of neutralizing antibodies and a focus of vaccine efforts. Here we report the structure at 3.5-Å resolution for an HIV-1-Env trimer captured in a mature closed state by antibodies PGT122 and 35O22. This structure reveals the prefusion conformation of gp41, indicates rearrangements needed for fusion activation, and defines parameters of immune evasion and immune recognition. Prefusion gp41 encircles N- and C-terminal strands of gp120 with four helices that form a membrane-proximal collar, fastened by insertion of a fusion peptide-proximal methionine into a gp41-tryptophan clasp. Spike rearrangements required for entry likely involve opening the clasp and expelling the termini. N-linked glycosylation and sequence-variable regions cover the prefusion closed spike: we used chronic cohorts to map the prevalence and location of effective HIV-1-neutralizing responses, which were distinguished by their recognition of N-linked glycan and tolerance for epitope-sequence variation.

؉ T cell decline, viral load, and CD4 ؉ T cell count at 6 months postinfection but not at later time points, suggesting that early events set the stage for the development of breadth. However, in a multivariate analysis, CD4 decline was the major driver of this association, as viral load was not an independent predictor of breadth. Mapping of the epitopes targeted by cross-neutralizing antibodies revealed that in one individual these antibodies recognized the membrane-proximal external region (MPER), while in two other individuals, cross-neutralizing activity was adsorbed by monomeric gp120 and targeted epitopes that involved the N-linked glycan at position 332 in the C3 region. Serum antibodies from the other four participants targeted quaternary epitopes, at least 2 of which were PG9/16-like and depended on the N160 and/or L165 residue in the V2 region. These data indicate that fewer than 20% of HIV-1 subtype C-infected individuals develop antibodies with cross-neutralizing activity after 3 years of infection and that these antibodies target different regions of the HIV-1 envelope, including as yet uncharacterized epitopes.

Neutralizing antibodies are likely to play a crucial part in a preventative HIV-1 vaccine. Although efforts to elicit broadly cross-neutralizing (BCN) antibodies by vaccination have been unsuccessful1-3, a minority of individuals naturally develop these antibodies after many years of infection4-7. How such antibodies arise, and the role of viral evolution in shaping these responses, is unknown. Here we show, in two HIV-1–infected individuals who developed BCN antibodies targeting the glycan at Asn332 on the gp120 envelope, that this glycan was absent on the initial infecting virus. However, this BCN epitope evolved within 6 months, through immune escape from earlier strain-specific antibodies that resulted in a shift of a glycan to position 332. Both viruses that lacked the glycan at amino acid 332 were resistant to the Asn332-dependent BCN monoclonal antibody PGT128 (ref. 8), whereas escaped variants that acquired this glycan were sensitive. Analysis of large sequence and neutralization data sets showed the 332 glycan to be significantly underrepresented in transmitted subtype C viruses compared to chronic viruses, with the absence of this glycan corresponding with resistance to PGT128. These findings highlight the dynamic interplay between early antibodies and viral escape in driving the evolution of conserved BCN antibody epitopes.

Identifying the targets of broadly neutralizing antibodies to HIV-1 and understanding how these antibodies develop remain important goals in the quest to rationally develop an HIV-1 vaccine. We previously identified a participant in the CAPRISA Acute Infection Cohort (CAP257) whose plasma neutralized 84% of heterologous viruses. In this study we showed that breadth in CAP257 was largely due to the sequential, transient appearance of three distinct broadly neutralizing antibody specificities spanning the first 4.5 years of infection. The first specificity targeted an epitope in the V2 region of gp120 that was also recognized by strain-specific antibodies 7 weeks earlier. Specificity for the autologous virus was determined largely by a rare N167 antigenic variant of V2, with viral escape to the more common D167 immunotype coinciding with the development of the first wave of broadly neutralizing antibodies. Escape from these broadly neutralizing V2 antibodies through deletion of the glycan at N160 was associated with exposure of an epitope in the CD4 binding site that became the target for a second wave of broadly neutralizing antibodies. Neutralization by these CD4 binding site antibodies was almost entirely dependent on the glycan at position N276. Early viral escape mutations in the CD4 binding site drove an increase in wave two neutralization breadth, as this second wave of heterologous neutralization matured to recognize multiple immunotypes within this site. The third wave targeted a quaternary epitope that did not overlap any of the four known sites of vulnerability on the HIV-1 envelope and remains undefined. Altogether this study showed that the human immune system is capable of generating multiple broadly neutralizing antibodies in response to a constantly evolving viral population that exposes new targets as a consequence of escape from earlier neutralizing antibodies.

A small proportion of HIV-infected individuals generate a neutralizing antibody (NAb) response of exceptional magnitude and breadth. A detailed analysis of the critical epitopes targeted by broadly neutralizing antibodies should help to define optimal targets for vaccine design. HIV-1-infected subjects with potent cross-reactive serum neutralizing antibodies were identified by assaying sera from 308 subjects against a multiclade panel of 12 "tier 2" viruses (4 each of subtypes A, B, and C). Various neutralizing epitope specificities were determined for the top 9 neutralizers, including clade A-, clade B-, clade C-, and clade A/C-infected donors, by using a comprehensive set of assays. In some subjects, neutralization breadth was mediated by two or more antibody specificities. Although antibodies to the gp41 membrane-proximal external region (MPER) were identified in some subjects, the subjects with the greatest neutralization breadth targeted gp120 epitopes, including the CD4 binding site, a glycan-containing quaternary epitope formed by the V2 and V3 loops, or an outer domain epitope containing a glycan at residue N332. The broadly reactive HIV-1 neutralization observed in some subjects is mediated by antibodies targeting several conserved regions on the HIV-1 envelope glycoprotein.