An effective HIV-1 vaccine will likely require the elicitation of broadly neutralizing antibodies (bNAb) to multiple epitopes such as the V2 apex, CD4 binding site (CD4bs), and Fusion Peptide (FP), though it is unclear how to best do so. Longitudinal studies of virus-antibody co-evolution in HIV-infected humans with bNAbs have illuminated many important aspects to consider in immunogen design, but inevitably represent one-off events. In this thesis, we characterized the virus-antibody co-evolution of a novel N160 glycan-independent V2 apex bNAb from an HIV-infected human and a CD4bs bNAb from a SHIV-infected rhesus macaque infected with a homologous envelope. During this co-evolution we observed transient deletions of CD4bs glycans in the viral quasispecies that preceded bNAb elicitation and suggested a potential strategy for priming CD4bs bNAbs in macaques more frequently. We incorporated that strategy into a glycan-hole SHIV infection model and observed significant increases in epitope targeting and CD4bs bNAb elicitation. We then translated this glycan de-shielding to generate SHIVs lacking glycans around the FP epitope. Infection with these SHIVs resulted in increases in FP targeting and FP bNAb elicitation. Together, these data highlight that glycan de-shielding is an effective, tunable method to immunofocus antibody responses to bNAb epitopes that may be employed in HIV vaccine design.