Search

HIV-1 virus entry into target cells requires the envelope glycoprotein (Env) to first bind the primary receptor, CD4 and subsequently the co-receptor. Antibody access to the co-receptor binding site (CoRbs) in the pre-receptor-engaged state, prior to cell attachment, remains poorly understood. Here, we have demonstrated that for tier-1 Envs, the CoRbs is directly accessible to full-length CD4-induced (CD4i) antibodies even before primary receptor engagement, indicating that on these Envs the CoRbs site is either preformed or can conformationally sample post-CD4-bound state. Tier-2 and tier-3 Envs, which are resistant to full-length CD4i antibody, are neutralized by m36.4, a lower molecular mass of CD4i-directed domain antibody. In some tier-2 and tier-3 Envs, CoRbs is accessible to m36.4 even prior to cellular attachment in an Env-specific manner independent of their tier category. These data suggest differential structural arrangements of CoRbs and varied masking of ligand access to the CoRbs in different Env isolates.

Efficient immune responses to HIV-1 gene products are essential elements to the development and design of an effective vaccine. Ideally, both humoral and cellular responses will be optimally elicited. It is therefore important to elucidate any factors that might limit the immunogenicity of HIV-1 proteins that are likely to be included in an effective vaccine. Since the HIV-1 exterior envelope glycoprotein gp120 is a major target for neutralizing antibodies, it is a virtual certainty that this gene product will be a component of any vaccine that seeks to elicit neutralizing antibody responses from the host humoral immune system. We report here the testing of several HIV-1 gp120 variants derived from a primary isolate that appears deficient in eliciting immune responses at both the level of CD4+ help and consequently in the generation of high-affinity IgG antibody responses in small animals. Factors limiting an effective immune response include (a) envelope glycoprotein strain variation decreasing functional T-cell help, (b) alteration of the glycosylation patterns of gp120 by expression in different cell types, and (c) the native structure of gp120 itself, which may limit the elicitation of effective T-cell help during natural infection or during parenteral immunization in adjuvant. Such limiting factors and others should be considered in the design and testing of gp120-based immunogens in small animals and possibly in primates as well.

Covalent conjugation of immune-stimulatory compounds to protein antigens is a potential means to self-adjuvant non-replicating subunit vaccines. Previously, it was demonstrated that covalent coupling of a Toll-like receptor (TLR) ligand to the exterior HIV-1 envelope glycoprotein, gp120, enhanced its immunogenicity. However, the consequences of chemical conjugation to gp120 on broadly neutralizing antibody (bNAb) epitopes were so far not examined. Here, we conjugated a TLR7/8 ligand to lysine residues on gp120 using NHS-PEO8-maleimide linkers and investigated if this affected Ab recognition of the CD4 binding site (CD4bs), a highly conserved target for bNAbs. We demonstrate that the recognition of the CD4bs was reduced following coupling, especially at a higher coupling ratio. These results have implications for the coupling of ligands to vaccine antigens where elicitation of humoral immune responses to specific neutralizing determinants is desired.

The HIV-1 exterior envelope glycoprotein gp120 binds receptor (CD4) and co-receptors (CCR5/CXCR4) and is a major target for neutralizing antibodies. The two functionally conserved regions of gp120 involved in receptor binding are conformational in nature. It is likely that the elicitation of neutralizing antibodies to these targets will benefit by presentation of these sites to the humoral immune system under physiologic conditions. Initially, we investigated the ability of the molecular adjuvant C3d to enhance antibody responses to variant gp120 glycoproteins in phosphate-buffered saline (PBS). We utilized a gp120 variant glycoprotein deleted of N- and C-terminal sequences (gp120DeltaC1/C5) originally designed to eliminate immunodominant, non-neutralizing epitopes and characterized this protein when fused to two C3d elements (gp120DeltaC1/C5(C3d)(2)). In PBS, the gp120DeltaC1/C5(C3d)(2) proteins are able to elicit gp120 binding antibodies more efficiently than gp120 lacking C3d moieties. We then asked if we could observe C3d-enhanced immunogenicity of gp120 in the presence of the classical oil-in-water adjuvant, Ribi. In the presence of the Ribi, which contains the TLR-4 agonist monophospholipid A (MPL), antibodies elicited by the gp120DeltaC1/C5(C3d)(2) were of higher titer than those elicited by the identical protein in PBS. To determine if the elicited secondary response was due to a synergy between the C3d repeats and the Ribi, we then inoculated gp120DeltaC1/C5 protein in Ribi and observed that similar titers of anti-gp120 antibodies were elicited in comparison to the gp120DeltaC1/C5(C3d)(2) protein also inoculated in Ribi adjuvant. In Ribi, there was a small but consistent increase in gp120-specific antibody titer of a gp120DeltaC1/C5(C3d)(2) prime followed by two gp120DeltaC1/C5 boosts compared to three inoculations of either the gp120DeltaC1/C5 proteins or the gp120DeltaC1/C5(C3d)(2) proteins alone. We conclude that the molecular adjuvant C3d demonstrates utility in conditions where physiologic presentation of native protein structures is desired, but may have less benefit in the context of a relatively potent protein adjuvant such as Ribi.

A critical component of an effective HIV vaccine will be the induction of broadly neutralizing antibodies. Comprising the HIV spike, the exterior envelope glycoprotein gp120 and the transmembrane glycoprotein gp41 mediate receptor binding, viral entry, and are the targets for neutralizing antibodies. The gp120 and gp41 glycoproteins are derived from the gp160 precursor glycoprotein and following gp160 glycosylation, oligomerization and cleavage in the endoplasmic reticulum and Golgi, remain as non-covalently associated trimers of heterodimers. Previously, using cell-surface envelope glycoproteins derived from infection of a laboratory-adapted HIV-1 strain, a correlation had been established between the binding of gp120-directed antibodies to the viral glycoprotein and the ability of the antibodies to neutralize laboratory-adapted isolates. However, this has been more difficult to demonstrate for glycoproteins derived from primary patient isolates. Here, using a FACS-based method, we report that only gp120-directed neutralizing antibodies and the neutralizing ligand soluble CD4 efficiently bind to glycoproteins derived from the JR-FL primary isolate provided that the gp160 precursor protein is efficiently cleaved. Precursor cleavage was demonstrated by cell-surface biotinylation and Western blotting. In stark contrast, both non-neutralizing and neutralizing antibodies bind non-cleaved envelope glycoproteins from JR-FL and YU2 isolates. These data imply that significant changes in Env spike structure are dependent upon precursor gp160 cleavage and are consistent with a restricted-binding-to-Env model of neutralization. The data also have implications in regards to the use and design of non-cleaved envelope glycoprotein trimeric immunogens as a means to selectively and preferentially present neutralizing epitopes to the host immune system.

The elicitation of broadly neutralizing antibodies directed against the human immunodeficiency virus type 1 (HIV-1) envelope glycoproteins, gp120 and gp41, remains a major challenge. Attempts to utilize monomeric gp120 as an immunogen to elicit high titers of neutralizing antibodies have been disappointing. Envelope glycoprotein constructs that better reflect the trimeric structure of the functional envelope spike have exhibited improved immunogenicity compared with monomeric gp120. We have described soluble gp140 ectodomain constructs with a heterologous trimerization motif; these have previously been shown to elicit antibodies in mice that were able to neutralize a number of HIV-1 isolates, among them primary isolate viruses. Recently, solid-phase proteoliposomes retaining the envelope glycoproteins as trimeric spikes in a physiologic membrane setting have been described. Here, we compare the immunogenic properties of these two trimeric envelope glycoprotein formulations and monomeric gp120 in rabbits. Both trimeric envelope glycoprotein preparations generated neutralizing antibodies more effectively than gp120. In contrast to monomeric gp120, the trimeric envelope glycoproteins elicited neutralizing antibodies with some breadth of neutralization. Furthermore, repeated boosting with the soluble trimeric formulations resulted in an increase in potency that allowed neutralization of a subset of neutralization-resistant HIV-1 primary isolates. We demonstrate that the neutralization is concentration-dependent, is mediated by serum IgG and that the major portion of the neutralizing activity is not directed against the gp120 V3 loop. Thus, mimics of the trimeric envelope glycoprotein spike described here elicit HIV-1-neutralizing antibodies that could contribute to a protective immune response and provide platforms for further modifications to improve the efficiency of this process.

The human immunodeficiency virus (HIV-1) exterior envelope glycoprotein, gp120, mediates receptor binding and is the major target for neutralizing antibodies. Primary HIV-1 isolates are characteristically more resistant to broadly neutralizing antibodies, although the structural basis for this resistance remains obscure. Most broadly neutralizing antibodies are directed against functionally conserved gp120 regions involved in binding to either the primary virus receptor, CD4, or the viral coreceptor molecules that normally function as chemokine receptors. These antibodies are known as CD4 binding site (CD4BS) and CD4-induced (CD4i) antibodies, respectively. Inspection of the gp120 crystal structure reveals that although the receptor-binding regions lack glycosylation, sugar moieties lie proximal to both receptor-binding sites on gp120 and thus in proximity to both the CD4BS and the CD4i epitopes. In this study, guided by the X-ray crystal structure of gp120, we deleted four N-linked glycosylation sites that flank the receptor-binding regions. We examined the effects of selected changes on the sensitivity of two prototypic HIV-1 primary isolates to neutralization by antibodies. Surprisingly, removal of a single N-linked glycosylation site at the base of the gp120 third variable region (V3 loop) increased the sensitivity of the primary viruses to neutralization by CD4BS antibodies. Envelope glycoprotein oligomers on the cell surface derived from the V3 glycan-deficient virus were better recognized by a CD4BS antibody and a V3 loop antibody than were the wild-type glycoproteins. Absence of all four glycosylation sites rendered a primary isolate sensitive to CD4i antibody-mediated neutralization. Thus, carbohydrates that flank receptor-binding regions on gp120 protect primary HIV-1 isolates from antibody-mediated neutralization.

The binding of HIV-derived recombinant soluble (s)gp120 to the CD4+/CXCR4+A3.01 T cell line inhibits the binding of the CXCR4-specific monoclonal antibodies 12G5, which interacts with the second extracellular loop, and 6H8, which binds the NH2terminus. We have used this as an assay to analyse the interaction of recombinant sgp120 from diverse viral origins with CXCR4. The strength of the interaction between sgp120 and CXCR4 correlated with sgp120 affinity for the CD4–CXCR4 complex, and the interaction of sgp120MNand sgp120IIIBwith CXCR4 was highly dependent on the level of CD4 expressed on a variety of different T cell lines. sgp120 from X4, R5X4, and R5 viruses interacted with CXCR4, although the R5 sgp120–CXCR4 interactions were weaker than those of the other gp120s. The interaction of sgp120IIIBor sgp120MNwith CXCR4 was inhibited by neutralizing monoclonal antibodies that prevent the sgp120-CD4 interaction but also by antibodies specific for the gp120 V2 and V3 loops, the CD4-induced epitope and the 2G12 epitope, which interfere weakly or not at all with CD4-sgp120 binding. The binding to A3.01 cells of wild-type sgp120HxB2, but not of sgp120 deleted in the COOH and NH2termini, interfered with 12G5 binding in a dose-dependent manner. Further deletion of the V1 and V2 loops restored CXCR4 binding activity, but additional removal of the V3 loop eliminated the gp120-CXCR4 interaction, without decreasing the affinity between mutated sgp120 and CD4. Taken together, these results demonstrate that the interactions between sgp120 and CXCR4 are globally similar to those previously observed between sgp120 and CCR5, with some apparent differences in the strength of the sgp120-CXCR4 interactions and their dependence on CD4.