A retrospective computational validation of a clinically evaluated recombinant envelope protein tetravalent dengue vaccine

Dengue virus (DENV), a member of the Orthoflavivirus genus, remains a major global health concern due to its high prevalence and the risk of severe disease. Vaccine development is challenged by the presence of four antigenically distinct serotypes (DENV1–4), as secondary infections with a heterologous serotype can result in antibody-dependent enhancement (ADE). Licensed vaccines like Dengvaxia and Qdenga rely on yellow fever virus (YFV)-based vectors encoding DENV structural proteins, though ADE risks persist. An alternative subunit approach, TetraVax-DV (V180), incorporates truncated envelope (Env) proteins from all four serotypes and is under development as a heterologous booster following YFV-DENV priming. Clinical data show that V180 induces strong neutralizing antibody titers and memory B cell responses. In this study, we applied classical in silico tools to assess the structural and immunological profiles of V180 subunits and compared them to EDIII and r2ED, two DENV antigens previously tested in murine models. All antigens displayed favorable predicted solubility, antigenicity, and structural stability. Epitope mapping identified high-affinity T-cell epitopes across multiple HLA alleles, with r2ED and EDIII showing lower epitope density. Simulated immune kinetics predicted robust antibody production, memory B cell formation, Th1 polarization, and CD8⁺ T-cell activation, particularly for V180 and r2ED. However, limitations in predictive accuracy were noted, such as the erroneous docking of EDE1-C10 to EDIII, highlighting the need for improved modeling approaches incorporating structural dynamics and artificial intelligence (AI) tools. These results support further development of V180 and r2ED, while emphasizing the strengths and boundaries of computational vaccinology.