Human immunodeficiency virus (HIV) is a globally devastating virus for which no effective vaccine currently exists, despite ongoing research efforts to create one. A significant arm of HIV research concentrates on creating a vaccine that induces broadly neutralizing antibodies (bNAbs), or antibodies that can bind and neutralize a large range of viruses. Although bNAbs have been isolated from HIV-1-infected patients, it is difficult to generate these antibodies through vaccination because the germline precursors of the antibodies do not bind HIV antigens. To better understand how to generate bNAbs through immunization, a recent eLife publication headed by Andrew Borst (Veesler Lab, UW) and Connor Weidle (Pancera Group: Stamatatos lab, Vaccine and Infectious Disease Division) characterized germline precursor binding of VRC01 to various 426c clade HIV antigens.
The HIV envelope (Env) glycoprotein is important for attachment and fusion to host cell membranes and contains the CD4 binding site, a conserved site that is recognized by neutralizing antibodies. Although HIV infection elicits antibodies that neutralize virus, rapid mutations acquired during viral replication allow HIV to evade and quickly outpace the immune system, as changes in Env preclude neutralization by previously effective antibodies. Therefore, an effective vaccine must target epitopes that are shared among many HIV viruses. Importantly, a class of bNAbs called VRC01 have been isolated from multiple HIV-infected patients and have been shown to successfully neutralize a broad range of HIV-1 strains by targeting the receptor binding site epitope. However, effective VRC01 bNAbs require years of HIV exposure to reach maturity, as antibodies hone their affinity for the antigen over time. Unlike mature VRC01 antibodies, their inferred germline precursors are unable to bind and neutralize HIV, hindering experimental induction of effective bNAbs by HIV immunogens. The binding of germline VRC01 bNAbs to HIV is thought to be sterically inhibited by N-linked glycosylation sites (NLGS) present at various positions on the surface on Env, specifically glycans at position N276. Removal of these glycans restores binding and neutralization ability of the germline antibody precursors; however, the modified HIV constructs fail to elicit productive bNAbs, limiting their usefulness as immunogens for vaccine research.
To interrogate how germline antibody engagement with HIV vaccine immunogens might be facilitated, the authors characterized the binding of VRC01 germline antibody with two of the lab’s engineered clade C HIV immunogens: a 426c native-like soluble trimer that lacks NLGSs at three positions, including N276, and a monomeric 426c core construct that contains all wildtype NLGSs. Using an engineered disulfide bond to stabilize the interaction for structural analysis, the authors purified the complex and used cryo-electron microscopy (cryoEM) to visualize the contact between the native-like trimer and the VRC01 germline antibody. The structure showed for the first time the interactions between native-like Env-derived immunogens with a germline precursor of VRC01-class antibodies and revealed that the naturally missing glycans at position N234 and N362 could be increasing accessibility to the site where bnAbs bind, possibly explaining the ability of this particular construct to bind germline VRC01.
Although engineered constructs that lack glycans are able to bind germline VRC01-class antibodies, this interaction does not mimic natural infections, as NLGSs are present on circulating HIV strains. Using biolayer interferometry (a measure of molecule interaction determined by light absorbance), crystal structure analysis, and mass spectrometry the authors were able to show for the first time that a wild type 426c core construct which retains short glycans at N276 (Man5) can bind germline VRC01-class antibodies, unlike what has been seen with other immunogens. The authors speculate that this successful binding may be due to the 426c core’s natural absence of glycans at N234 and N362, allowing for improved interaction between the antibodies and their epitope. Together, these results demonstrated that germline VRC01-class antibodies may binding 426c core in both the absence and presence of NLGS at critical positions, something that has not previously been shown in germline VRC01 with a protein containing a short glycan at N276, suggesting that HIV proteins with shorter glycoforms could potentially be used to engage these germline antibodies.
In this study, the authors characterize how immunogens that better mimic HIV virus could be used to target germline VRC01-class antibodies in order to elicit bnAbs capable of providing protection from HIV infection. Stamatatos explained that “our data indicate that germline VRC01-class antibodies can bind their epitope, even in the presence of short glycans expressed on the N276 NLGS,” a feature that previously has been “thought to present a major obstacle for the recognition of HIV-1 Env by the germline forms of VRC01-class antibodies.” Stamatatos said these results introduce the question of whether VRC01 bnAb development is initiated by germline antibody binding with HIV Env that have glycans at position N276 and whether recombinant Envs that express these same glycans can be used to prime antibodies to become VRC01 bnAbs.
Borst AJ, Weidle CE, Gray MD, Frenz B, Snijder J, Joyce MG, Georgiev IS, Stewart-Jones GB, Kwong PD, McGuire AT, DiMaio F, Stamatatos L, Pancera M, Veesler D. 2018. Germline VRC01 antibody recognition of a modified clade C HIV-1 envelope trimer and a glycosylated HIV-1 gp120 core. eLife. Nov 7;7. pii: e37688. doi: 10.7554/eLife.37688.
This work was funded by the National Institute of General Medical Sciences, the National Institute of Allergy and Infectious Diseases, Pew Biomedical Scholars Award, Investigators in the Pathogenesis of Infectious Disease Award, the Netherlands Organization for Scientific Research, and the European Molecular Biology Organization.
Fred Hutch/UW Cancer Consortium members Leonidas Stamatatos (UW/FH) and Andrew McGuire (FH) contributed to this work.