In this study, we have produced soluble versions of Env glycoproteins derived from a primary isolate stabilized by Phe 43 pocket filling mutations and inter-domain cysteine bonds. We have tested these glycoproteins for immunogenicity in rabbits as a prime for soluble gp140 molecule boost. Breadth of neutralization indicate that these glycoproteins significantly increase neutralization breadth as a prime for the soluble gp140 trimer boosts and both are interesting platforms for further modification to better elicit broadly neutralizing antibodies.? ? We have continued on to make cysteine-stabilized versions of the core proteins? ? Stabilization of gp120 into the CD4-bound Conformation and Immunofocusing? ? As a long-term collaborative effort, using the initial core-CD4 crystal structure, we have attempted to stabilize gp120 into the CD4-bound conformation to alter its immunogenicity. The rationale flows primarily from the structure, which allows precise cysteine-pair-based mutagenesis to conformationally fix gp120, and from ligand micocalorimetry experiments that demonstrate that free gp120 is an extremely flexible molecule 11. We have generated a series of gp120 cores that contain up to 4 pairs of cysteines and confirmed bond formation by structure. In my lab, we have performed microcalorimetry demonstrating that the molecules become progressively more fixed by the introduction of the cysteine pairs. Because the cores present unnatural surfaces to the immune system (due to loop and N- and C-terminal truncations and the lack of gp41 association), we have employed a core prime: trimer boost immunogenicty strategy to immunofocus the antibody response to the conserved elements between the stabilized cores and soluble trimers. Studies are under evaluation comparing the immunogenicity of unmodified core proteins to those containing partially stabilizing cavity filling mutations, to those containing one to four additional cysteine pairs. To date, the neutralizing activity elicited by the prime:boost immunofocusing generated enhanced neutralization breadth that correlates with increasing molecular stabilization by introduction of the additional cysteine pairs. These exciting preliminary observations will be re-tested and confirmed within the next one to two years. At this juncture, we are cautiously optimistic about eliciting neutralizing antibodies by the CD4-stabilized gp120 conformation as a line of investigation. Due to the preliminary data of enhanced neutralization breadth, we anticipate that neutralizing antibodies directed against the CD4 binding site have been elicited. To confirm this, we will use the comparative adsorptions of WT or CD4 binding deficient gp120 glycoproteins coupled to solid phase beads described in Project 2. Although these initial results are encouraging, at a minimum, we need to increase the potency of this neutralizing response. ? Another form of immunofocusing is to modify the core gp120 (or trimeric Env) by glycan masking, leaving primarily the CD4 binding site/b12 surface non-glycosylated as an immunogenic surface to be recognized by B cells in the host repertoire. In the ideal case, when B cells scan the surface for non-self protein determinants, only this region will be accessible. The B cell repertoire specific for the CD4 binding patch will probably be quite low in frequency, so the antibody responses will likely be low initially, unless repeated boosts with glycosylated immunogens are performed. We will likely need to add heterologous T help to the heavily glycosylated immunogens. We will systematically address if we will also need to add B cell or DC targeting moieties (i.e., C3d or a DEC single chain antibody) or TLR ligands to optimize immunogenicity. We have observed that the heavily glycosylated trimeric molecules that we inoculated in collaboration with Dennis Burton did become less immunogenic, however to date we have not observed enhanced neutralization. But glycan coating is just the start of the process as mentioned. Other glycan masking, immunofocusing approaches will include adding on glycans to the YU2 foldon trimers in V1 and V3 where we mapped a significant fraction of the trimer-elicited, strain-restricted responses 12.? ? b) Use of the New and Novel F105- and b12-gp120 Structures to Guide Immunogen Design? Another extremely exciting line of investigation is to use the difference maps between the two broadly neutralizing ligands b12 and CD4 and the similar but non-neutralizing CD4 binding site antibody F105 to impact on immunogen design. The new structures of gp120 in complex with b12 and F105 are now being defined in the Kwong lab in collaboration with the Gary Nabels laboratory. Somewhat surprisingly, the footprints and angles of approach between the neutralizing and non-neutralizing ligands are not that dissimilar, but by utilizing the details of the atomic structures, it may be possible to place glycans to occlude the binding of F105 and most non-neutralizing CD4 binding site antibodies, while permitting binding by b12-like or CD4-like antibodies. In a parallel effort with Bill Schief, we will try to stabilize the core in the CD4 or b12 bound conformation while eliminating recognition by the non-neutralizing class of CD4 binding site antibodies typified by F105. The Kwong and Nabel labs are also trying to stabilize the b12 binding site present on the outer domain, and if successful, this immunogen can be utilized in prime:boost strategies with trimers or other candidate immunogens to enhance immunofocusing on the desired neutralizing surfaces.? ? c) Exploiting Adjuvants, Innate and Adaptive Immunity to Enhance Env Immunogenicity ? For selected model immunogens, we will assess ways to quantitatively or qualitatively enhance antibody responses, while also eliciting cellular responses. The recent discovery that signaling through particular TLRs can generate both antibody and cellular responses specific for HIV proteins has garnered much excitement 14, 15. Such cross-priming might permit a protein immunogen to interface with the DNA-adeno platform in vaccine trials at the VRC to boost both antibody and cellular responses. Accordingly, my laboratory, in collaboration with Robert Seder at the VRC and Gunilla Karlsson Hedestam at the Karolinska Research Institute, will assess novel ways to couple TLR ligands to candidate gp120 molecules without altering the antigenic surface of the glycoprotein. Besides the direct conjugation of selected TLR ligands, we are exploring other novel, less invasive ways to link TLR ligands to gp120 candidate proteins. We will assess coupling the TLR ligands CpGs or polyI:C to a protamine tail appended to gp120. We are producing gp120 as a fusion protein with the TLR-5 ligand, flagellin. And in collaboration with Amos Smith at UPenn, we will attach a C-terminal unpaired cysteine residue for sulfhydral conjugation with selected TLR ligands. Once generated, these conjugates will be characterized to determine TLR ligand stoichiometry requirements, tested for functional activity using reporter cell lines and tested for immunogenicity relative to controls lacking the TLR ligands or possessing soluble TLR ligands. Combinations of the gp120-conjugated TLR ligands can be tested for synergy in the elicitation of B cell antibody responses that might in turn increase the levels of circulating antibodies.? In sum, this integrated approach: to design immunogens based upon structure and biochemistry that selectively present neutralizing determinants to the immune system, then to optimize the immunogenicity by the addition of TLR ligands, other defined molecular adjuvants or improved classical adjuvants is an overall approach to be exploited over the next several years of vaccine development in my laboratory
Showing the most recent 10 out of 14 publications
