Reevaluating Human Cytomegalovirus Vaccine Design

Vaccines to prevent HCMV infection have primarily included viral antigens identified as key targets of neutralizing antibody (NAb) and/or T cell responses in healthy adults with chronic HCMV infection, such as glycoprotein B (gB), the glycoprotein H-anchored pentamer complex (PC), and the unique long 83 (UL83)-encoded phosphoprotein 65 (pp65). However, the protracted absence of a licensed HCMV vaccine that reduces the risk of infection in pregnancy regardless of serostatus warrants a systematic reassessment of assumptions informing vaccine design. To illustrate this approach, we considered the hypothesis that HCMV proteins infrequently detected as targets of T cell responses may be important vaccine antigens. We used an extant dataset from a T cell profiling study to test whether HCMV proteins recognized by only a small minority of participants encompass any T cell epitopes. Our analyses demonstrate a prominent skewing of T cell responses away from most viral proteins – although they contain robust predicted CD8 T cell epitopes – in favor of a more restricted set of proteins. Our findings raise the possibility that HCMV derives a net benefit from avoiding T cell recognition of certain proteins, and that contrary to current designs approaches, including them as vaccine antigens could effectively take advantage of this vulnerability