HIV envelope glycoprotein gp120 contains epitopes which are targeted by broadly cross-reactive neutralizing antibodies in humans and which depend on the native protein conformation. Monoclonal antibodies to these sites can neutralize a broad range of HIV-1 isolates and have protected monkeys against viral challenge by iv and oral routes. Based on mapping these epitopes on the three dimensional structure of gp120, we have found a structurally independent site, where foreign protein sequences can be linked to gp120 without disrupting the native folding of the neutralizing sites. By linking gp120 to a carrier protein capable of self-assembly, we could enhance the intrinsic vaccine potency of gp120 by assembling virus-like particles, while retaining important conformational sites needed to elicit these antibodies. Two carrier proteins were inserted at this site, hepatitis B surface antigen (HBsAg) and hepatitis B core antigen. Both carriers share the ability to form particles, but core antigen forms rigid regular structures, while surface antigen forms flexible lipo-protein micelles. When expressed in tandem with gp120, core-gp120 hybrids assembled poorly, while HBsAg-gp120 hybrids assembled particles efficiently. Most importantly, gp120 in the hybrids showed normal glycosylation, high affinity binding of the natural receptor CD4, and bound a panel of broadly reactive human neutralizing monoclonals, indicating that it was folded correctly in the native conformation. The particulate form of gp120 banded in CsCl at a light density, suggesting a lipid content of about 10 to 20%. Electron microscopy showed 25 nm particles, similar to surface antigen alone. By analogy with HBsAg, these particles diplay an array of about 200 gp120 molecules at a lipid/water interface, closely resembling the surface of HIV virions. Assembly of gp120 into virus-like particles has been a long-term goal of this laboratory. In the case of other successful vaccines, such as HBsAg for hepatitis B virus, and L1 capsid protein of human papilloma virus, assembled particles were up to 1,000-fold more potent than the same weight of monomeric proteins. In order to test the immunogenicity of our particles, we had to modify our methods of expression and purification to increase our yield of particles. We have successfully converted from expression in vacccinia recombinants to expression in baculovirus, with the ability to produce hundreds of micrograms needed for immunological studies. The baculo expressed hybrids assemble as before, have normal glycosylation, and bind neutralizing monoclonal antibodies. On EM, they are 30 nm particles. We have expressed three envelope types: IIIB, 89.6, and SIV453 in baculovirus. The first two correspond to SHIV challenge strains. The last one could be used to protect monkeys from a potent SIV challenge. This is the first multimeric gp120 vaccine, and we anticipate that it may elicit a strong immune response in mice, rabbits and monkeys. If monkeys do produce high titered neutralizing antibodies, they will be challenged with titered stocks of SHIV virus of the same or different envelope type. In addition, an SIV challenge will be attempted, which may more closely represent a natural HIV infection with a highly adapted virus. Through potent carrier and multimer effects, particulate forms of gp120 and other viral antigens may have a profound effect on HIV immunology. This project is funded by an intramural NIH grant. Similar particles have been made of VEE envelope glycoproteins E1 and E2, to test whether particle formation can be generalized to other enveloped viruses.
