Adding to the Armamentarium of Broadly Neutralizing Antibodies
Broadly neutralizing antibodies (bNAbs) to the HIV Envelope spike are considered valuable tools for guiding the design of immunogens for AIDS vaccine candidates developed to induce such antibodies. Only a handful of bNAbs—including b12, 2G12, 4E10 and 2F5—were available, until last September, when IAVI researchers, in collaboration with researchers from The Scripps Research Institute in La Jolla, California, reported the isolation of two new bNAbs called PG9 and PG16 (1). Then, at the AIDS Vaccine 2009 conference in October, researchers from the Vaccine Research Center (VRC) at the US National Institute of Allergy and Infectious Diseases (NIAID) reported the isolation of three new bNAbs, including one called VRC01 (see Raft of Results Energizes Researchers, IAVI Report, Sep.-Oct. 2009).
Now, a group led by Antonio Lanzavecchia at the Institute for Research in Biomedicine (IRB) in Switzerland, has identified three additional bNAbs from HIV-infected individuals (2). “Between the three groups, we will have doubled the number of useful antibodies,” says Robin Weiss, a professor emeritus of viral oncology at University College London and one of the researchers involved in this study.
None of the three bNAbs identified in the PLoS One study “has the potency and breadth combined in a single antibody that say PG9 and PG16 show or that VRC01 shows,” Weiss says. Each of the three antibodies, called HJ16, HGN194, and HK20, recognizes a different part of the Env spike—the CD4 binding site, the tip of the V3 loop, and an epitope on the HR-1 region of gp41, respectively. This suggests “that there are a number of ways in which to neutralize HIV,” says Peter Kwong, chief of the structural biology section at the VRC, who was not involved in the PLoS One study.
To find the antibodies, the researchers tested sera from about 400 individuals infected with HIV clades prevalent in Africa, such as A and C, for their ability to neutralize a panel of viruses from different clades. They then used Epstein-Barr virus to immortalize the memory B cells of 21 donors with a broad neutralization profile to isolate monoclonal antibodies (mAbs). When they measured binding of these antibodies to recombinant Env proteins, including trimeric gp140 (which contains all protein parts of the Env spike), monomeric gp120, and gp41, they found that 58 mAbs bound to at least one recombinant Env protein from different clades, including clades A and C.
They then found that more than half of the 58 mAbs neutralized at least one HIV isolate. However, only three were broadly neutralizing, in that they neutralized virus from at least two clades and more than one virus per clade. This suggests that the ability of patient sera to broadly neutralize HIV may often come from the combined effects of antibodies that are not broadly neutralizing in and of themselves, and only rarely from a few very potent antibodies. “The main message is that there is an extensive number of neutralizing antibodies that can be retrieved from the human memory B cells of infected people, but the breadth of neutralization of these antibodies is very limited,” adds Davide Corti, the study’s first author, who is now director of the antibody discovery unit at Humabs, a spin-off company of the IRB.
Of the three bNAbs, HJ16 is “perhaps the most interesting,” Weiss says. Like b12, it binds the CD4 binding site of gp120, although to a different part. While its breadth is similar to b12, it best neutralizes those strains that b12 does not neutralize. “If you put the two together you get fantastic neutralization,” says Weiss. “[This] is against the dogma that in general monoclonal antibodies neutralize first of all the [neutralization] sensitive viruses and after that some resistant viruses,” Corti says. “It is indeed unusual.”
With the identification of HJ16, there are now three broadly neutralizing antibodies—b12, VRC01, and HJ16—that recognize a different part of the CD4 binding site, says Dennis Burton, a professor of immunology and microbial science at The Scripps Research Institute, who was involved in the PG9/16 study. “It adds to our knowledge about the CD4 binding site and it will help us to design [vaccine] candidates based around the CD4-binding site,” Burton says.
The second antibody, HGN194, recognizes a very conserved epitope, Corti says, in the V3 loop of gp120, which interacts with the CCR5 coreceptor during the HIV entry process. The V3 epitope consists of a continuous stretch of amino acids, which may make immunogen design for this type of antibody specificity easier, he adds.
The third antibody, HK20, binds to a region of gp41 that was thought to be inaccessible because it is close to the viral membrane and only exposed transiently, just before fusion takes place. This is probably why HK20 only becomes a potent and broadly neutralizing antibody as an Fab fragment, says Corti.
The method used to isolate PG9/16 involved screening for neutralization first. But Corti and colleagues identified their crop of bNAbs by first screening for their ability to bind recombinant Env protein and only later for their ability to neutralize HIV. “We now think that that’s not the smartest way to do it,” Weiss says, because an approach that first screens for binding might miss antibodies that can neutralize HIV even though they don’t bind well to recombinant Env proteins. His group is now screening for neutralization first, Weiss says.“There are better robotic methods to pick up neutralizers, and so over the last year it’s become much easier to do screening by primary neutralization.” —Andreas von Bubnoff