One of the most popular and promising targets for HIV vaccine development are immunogens derived from gp120, an envelope glycoprotein essential for viral entry into CD4 cells. Recently, our collaborators have isolated multiple HIV broadly neutralizing antibodies to gp120 and shown that they require certain N-linked glycans for activity. However, some of these antibodies do not bind the free N-linked glycans, suggesting that at least some of them may also interact with the protein portion of gp120. The antibody.gp120 complexes too large for generation of high-resolution structures by NMR spectroscopy, and as the glycan appear to play key roles in the antibody epitopes, crystallization of the complexes is a daunting task. Computational modeling is an attractive approach for this problem, but purely computational approaches can generate models of questionable accuracy, and require empirical constraints or testing in order to generate a reliable model. We propose to characterize the various broadly neutralizing antibody epitopes using hydroxyl radical protein footprinting of the gp120-antibody complexes, a technique that labels a broad variety of amino acid side chains based on their accessibility to solvent. In order to improve the usefulness of the footprinting data for accurate, high-resolution model building, we propose to develop a number of improved footprinting methods, including accurate quantitation at the amino acid level to improve structural resolution and normalization protocols to generate absolute solvent accessible surface area values from footprinting data. We also propose to develop an appropriate scoring function to utilize solvent accessible surface areas as a constraint in molecular dynamics simulations, analogous to the use of distance constraints. From these improvements and their application to the characterization of gp120-antibody complexes, we anticipate the generation of accurate, experimentally-constrained models that correctly identify the epitope for each antibody. These models will be very important for the rational design of immunogens to raise the corresponding broadly neutralizing antibodies in a host through immunization.
HIV remains a devastating disease throughout the world, especially in developing countries. Despite the urgent need, development of a vaccine to HIV remains elusive. Recently, a series of antibodies that neutralize a broad variety of HIV serotypes through interaction with the envelope glycoprotein gp120 have been isolated from HIV-positive patients. These antibodies require the glycan to bind gp120, but at least some of them do not bind solely to the glycan. However, the epitope that the antibody recognizes has not been defined, so it is not possible to develop an immunogen to raise the antibody in a host through a vaccine. We propose to utilize hydroxyl radical footprinting coupled with computational modeling to characterize and model the complex between gp120 and each of the broadly neutralizing antibodies. These models will aid in the rational development of immunogens to formulate effective anti-HIV vaccines.
