With the exception of inactivated vaccines, all SARS-CoV-2 vaccines currently used for clinical application focus on the spike envelope glycoprotein as a virus-specific antigen. Compared to other SARS-CoV-2 genes, mutations in the spike protein gene are more rapidly selected and spread within the population, which carries the risk of impairing the efficacy of spike-based vaccines. It is unclear to what extent the loss of neutralizing antibody epitopes can be compensated by cellular immune responses, and whether the use of other SARS-CoV-2 antigens might cause a more diverse immune response and better long-term protection, particularly in light of the continued evolution towards new SARS-CoV-2 variants. To address this question, we explored immunogenicity and protective effects of adenoviral vectors encoding either the full-length spike protein (S), the nucleocapsid protein (N), the receptor binding domain (RBD) or a hybrid construct of RBD and the membrane protein (M) in a highly susceptible COVID-19 hamster model. All adenoviral vaccines provided life-saving protection against SARS-CoV-2-infection. The most efficient protection was achieved after exposure to full-length spike. However, the nucleocapsid protein, which triggered a robust T-cell response but did not facilitate the formation of neutralizing antibodies, controlled early virus replication efficiently and prevented severe pneumonia. Although the full-length spike protein is an excellent target for vaccines, it does not appear to be the only option for future vaccine design.
