HIV is a global health challenge, with more than 39.9 million people living with HIV in 2024. HIV-1 is a notoriously challenging vaccine target due to its rapid mutagenicity and dense and diverse glycan coating. Glycans are polymerized, branching carbohydrates which are post translationally attached to proteins at specific amino acid sites. Protein envelopes of pathogenic viruses such as Coronavirus, Ebola and Influenza are also glycosylated, however the HIV-1 virus envelope has the highest known glycan density which accounts for over ½ of the entire envelope molecular . The highly heterogenous glycans constantly shift over time to form a 3-Dimensional “shield” around the HIV envelope protein (Env) antigen which masks the virus from adaptive immune detection and response. This glycan shield is a significant hamper to vaccine design as it prevents broadly neutralizing antibodies from binding to the surface epitope. Experimental imaging methods like cryo-electron microscopy (cryo-EM) and nuclear magnetic resonance (NMR) also cannot capture high-resolution results of changing glycan shield temporal dynamics. This project applies a high-throughput computational pipeline developed by our group to atomistically simulate physiologically relevant 3D glycan clouds. We apply a combination of computational techniques including molecular dynamics simulations, homology and ab initio glycan modeling, to generate to examine how varying glycoform composition perturbs glycan coverage over the HIV Env protein epitope. These epitope shielding effects are quantified using previously designed tools of Glycan Encounter Factor and network analysis. Our results will identify vulnerabilities in the glycan shield by determining how individual glycan types and long-distance interactions across the shield impact epitope coverage. This work reduces experimental burden, informs targeted immunogen design with topological control over the glycan shield and is a translatable platform for an array of glycosylated viruses from Covid-19 to Dengue virus.
