A Boost for HIV Vaccine Design

See allHide authors and affiliations

Science  13 Aug 2010:
Vol. 329, Issue 5993, pp. 770-773
DOI: 10.1126/science.1194693

A major roadblock to the development of an effective vaccine against the human immunodeficiency virus (HIV-1) is the lack of an immunogen that elicits broadly protective antibodies (1). Passive transfer studies in animal models have associated protection with neutralizing antibodies and, encouragingly, serum studies show that a subset of HIV-infected individuals produces potent broadly neutralizing antibodies (2). Understanding the viral targets of such antibodies and how they achieve potent and broad neutralization has become a key endeavor in HIV vaccine research. On page 856 of this issue, Wu et al. (3) describe the isolation of particularly potent monoclonal broadly neutralizing antibodies using a novel selection strategy, and on page 811, Zhou et al. (4) solve the crystal structure of the most effective of these antibodies in complex with its target gp120, a viral envelope glycoprotein. These studies further invigorate the currently active field of discovering broadly neutralizing antibodies against HIV (2, 57) and provide valuable molecular information for rational vaccine design.

The isolation of monoclonal broadly neutralizing antibodies from HIV-infected donors has so far proven immensely difficult. Following their initial characterization in the 1990s, only a few were available by early 2009. Part of the difficulty is that in HIV infection, a vast number of antibodies are produced to the viral glycoproteins (gp120 and gp41) that do not bind to the native trimeric spike formed by these proteins (see the figure) and, hence, do not efficiently neutralize the virus. These antibodies generate considerable “noise” and present “needle-in-a-haystack” problems in screens based on recombinant gp120 and gp41 as selecting agents to capture immune cells (memory B cells) that produce antibodies of interest.

Neutralizing HIV antibodies.

Broadly neutralizing monoclonal antibodies (mAbs) target epitopes on the viral envelope spike at the surface of HIV-1. The spike is a heterotrimer containing the viral glycoproteins (gp120)3 (gp41)3. This model is generated by combining cryoelectron tomographic (13), crystallographic (14), and computational analyses. It shows the gp120 core structure (tan) fitted in the electron density map (gray) of the spike. The membrane proximal external region (MPER) and viral membrane are shown. The V1/V2 and V3 protein loops are represented as ovals (light green and light blue) at the top of the spike. Glycans (green and blue) are indicated. The model is derived from (5) with permission from Wolters Kluwer Health–Lippincott Williams and Wilkins; from (6) with permission from Elsevier; and from (13) with permission from Macmillan Publishers Ltd./Nature Publishing Group.


Walker et al. (8) circumvented this problem by selecting a donor with a high broadly neutralizing antibody response and then identified two antibodies (PG9 and PG16) with exceptional neutralizing breadth and potency through a large-scale direct functional screen of the individual's B cells. Corti et al. (9) used multiple viral envelopes in parallel capture assays to favor cross-reactive antibodies, and identified the broadly neutralizing antibody HJ16. Wu et al. used a donor with extraordinarily broad and potent serum neutralization directed mainly to the site in gp120 that binds to CD4 (CD4bs) (10), the receptor used by HIV to dock onto cells. Using an engineered gp120 that did not bind the great majority of nonneutralizing antibodies to gp120, but preferentially bound the broadly neutralizing antibody to the CD4bs (b12), Wu et al. selected B cells that produced potent neutralizing antibodies against the CD4bs of gp120. One of these antibodies, VRC01, neutralizes ∼90% of nearly 200 viruses of different HIV clades tested. The outstanding potency and breadth of VRC01, together with the findings of other recent studies (8, 11), show that one of the fears of this field—that potency increases only at the cost of reduced breadth of viral strains that can be neutralized—is unfounded. Another concern was that b12 might be a unique broadly neutralizing antibody to the CD4bs, and the CD4bs thus an almost intractable vaccine target. The description of several broadly neutralizing antibodies to the CD4bs from different HIV-infected individuals (3, 9) lays this concern to rest and promotes interest in designing immunogens to elicit antibodies to CD4bs.

Zhou et al. describe the high-resolution crystal structure of the antigen-binding fragment (Fab) of VRC01 in complex with a core gp120 molecule. The antibody shows some similarities to and differences from the CD4 receptor in recognizing the CD4bs on gp120. The antibody targets amino acid residues in the outer domain of gp120, whereas the CD4 receptor also targets a region known as the bridging sheet. This difference in recognition appears to render VRC01 a more potent neutralizer than a CD4-antibody hybrid molecule (an “immunoadhesin”), suggesting that the antibody has evolved the most favored angle of approach to the CD4bs and/or requires the least conformational perturbation to bind to its critical contact residues.

The challenge now is to translate understanding of potent broadly neutralizing antibodies such as VRC01 and PG9/PG16 into the design of immunogens that elicit these types of antibody when incorporated in an HIV vaccine. Attempts using the early, less potent broadly neutralizing antibodies have so far been unsuccessful but have stimulated the development of promising concepts such as “scaffolds,” in which viral epitopes that are targets for neutralization are presented in the context of proteins that lack some of the molecular trickery that HIV envelope proteins have evolved to evade antibody responses. The structure of the VRC01-gp120 complex provides detailed atomic-level data of a neutralizing epitope centered on the CD4bs and merits a massive effort in immunogen design.

Both arms of the adaptive immune response—cell-mediated immunity and humoral immunity mediated through antibodies—are likely to be important in developing a vaccine that will provide broad and long-lasting immunity to the diversity of HIV strains to which humans may be exposed. Cell-mediated immunity is important in controlling viral infection. Indeed, individuals with impaired cell-mediated immunity are prey to opportunistic viral infections such as herpes simplex virus, cytomegalovirus, and oncogenic viruses. But neutralizing antibodies are a good correlate of protection for successful antiviral vaccines such as yellow fever, smallpox, and measles (12). To prevent HIV from taking hold in the first place, humoral immunity in the form of broadly neutralizing antibodies in the blood and at mucosal surfaces are likely to be a key component of protection. Developments such as those reported by Wu et al. and by Zhou et al. will help identify immunogens that can elicit broadly neutralizing antibodies and thereby become an essential part of an AIDS vaccine.


View Abstract

Navigate This Article