Plasma from HIV-1–infected patients neutralizes strains of HIV-2 that have been incubated with soluble CD4 (open circles).

On page 1407, Decker and colleagues trick HIV into exposing itself. This trick revealed that a majority of infected individuals have broadly cross-reactive neutralizing antibodies against HIV in their circulation.

Antibodies that neutralize diverse strains of HIV have been difficult to find, despite decades of searching. Several gp120-specific neutralizing antibodies have been cloned. Some of them target the binding site for the HIV coreceptor C-C chemokine receptor 5 (CCR5), which is exposed only when gp120 engages the CD4 receptor on target cells. Many strains of HIV require CCR5 to infect cells, but this binding site was thought to be weakly antigenic as it is normally hidden from the immune system.

Decker and colleagues now prove this idea wrong and show that the coreceptor binding site is one of the most highly immunogenic regions of gp120. The authors incubated various strains of HIV-2 with soluble CD4 (sCD4) to induce exposure of the coreceptor binding site. They then tested whether plasma from HIV-1 infected individuals could neutralize the CD4-induced virus. To their surprise, they found that >90% of individuals infected with HIV-1 had antibodies that could neutralize highly divergent strains of HIV-2 and SIV, but only when the viruses were first treated with sCD4. This suggested that the coreceptor binding site is not only immunogenic, but also antigenically conserved among different HIV strains.

The authors suggest that antibodies against this well-concealed epitope may be elicited in vivo by shed gp120 proteins—prevalent during HIV infection—that bind to CD4-expressing cells causing the gp120 molecules to reveal their coreceptor binding sites to antibody-producing B cells. Thus, antibodies are generated, but are probably helpless to attack their real target: the still hidden sites on the gp120 of the intact virus.Whether it will be possible to take advantage of this antigenic site for vaccine development remains to be seen. Providing sCD4 in vivo is unlikely to work, as previous clinical trials showed that sCD4 actually enhanced virus replication at low concentrations. One possibility, the authors suggest, is to create a vaccine that elicits two kinds of antibodies: one that mimics CD4 and exposes the coreceptor binding site and another that binds the exposed site and prevents HIV from getting into cells.