Numerous HIV vaccine design strategies that have advanced towards early clinical trials include boosting immune response with a recombinant, truncated form of Env, the trimeric surface antigen on the HIV-1 virus. Production of this protein in quantities large enough to support clinical trials is a challenge, and vaccine development efforts now predominantly favor generation of the soluble, monomeric gp120 component of this protein. Thus far, the protein has not been effective at eliciting neutralizing antibodies in humans, and we hypothesized that a contributing cause for this lackluster performance was that the recombinant, truncated preparations of gp120 are not effective mimics of native Env. We have conducted glycosylation and disulfide bonding analyses of many gp120 and gp140 vaccine candidates destined for clinical trials and compared their molecular profiles to trimeric, native-like Env of multiple sequence types. The results show that a conserved “native” glycosylation profile exists among native-like Env trimers, but gp120 and uncleaved gp140s typically do not adopt this glycosylation profile. Furthermore, unless the truncated Env is expressed as a gp140 and engineered to stabilize the gp120/gp140 interface, such as by using the SOSIP modifications, its disulfide bonding profile is typically highly heterogenous, with the majority of the disulfide bonds matching a non-native profile. Taken together these results emphasize that many vaccine design and production efforts need reconsideration, if developers intend to deliver recombinant Env in a native conformation as part of their vaccine strategy.
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.