Central HIV-1 sequences are designed to reduce the distance between a vaccine reagent and diverse circulating strains, with the intent of maximizing the cross-reactive potential. The HIV RAD Team has made the first M group and B and C subtype central strains based on our design. An M group central Envelope protein is the most challenging, as it is the farthest from natural isolates and the most highly variable protein. Despite this, our first M group construct CON-6 produced a weakly viable CCR5-using pseudotyped virus;more importantly it retained important conformational attributes in terms of binding to a panel of key monoclonal antibodies, it was immunogenic, and had encouraging antigenic properties. These reagents were designed in 1999, and the global alignments can double in size regarding key subtypes, each year. Since our last submission, we have new encouraging preliminary results with a 2001 M group consensus Env called CON S, that stimulated cross-reactive neutralization responses against multiple non-B clade primary isolates. We intend to systematically explore whether further addition of the new data could improve reagent design. Also, phylogenetic theory has made recent progress, but the new methods are computationally intensive and can applied only to small data sets. We will design parallel implementations for supercomputers, to reanalyze HIV sequences with state-of-the-art methods. Preliminary data describing C subtype primary isolates indicate that isolates obtained soon after infection have distinct properties;we think it will be advantageous to concentrate on these viruses for central immunogens for vaccine design. Immune responses to the central vaccine reagents will be systematically compared to each other and to natural proteins from clades A, B and C in animal models. By comparing the breadth of response to individual antigens, both natural and central, we will ultimately be able to design a cocktail of immunogens that provides the greatest breadth;this requires first defining the response to each component.
Showing the most recent 10 out of 27 publications
