Optimizing Immunity and Maternal Host Defense Against Congenital Cytomegalovirus Infection

ABSTRACT Development of an effective preventative vaccine against human cytomegalovirus (HCMV) for women of childbearing age is a major public health priority. Subunit vaccines are in clinical trials, but uncertainty exists about optimal platforms and correlates of protection. Live, attenuated vaccines may induce a broader repertoire of immune responses, but striking the balance between safety and immunogenicity is challenging. The long-term goal of this ongoing collaboration is to use the guinea pig cytomegalovirus (GPCMV) model to test hypotheses about optimal design of safe and effective vaccines against congenital infection. We recently discovered that inclusion of a pentameric complex (PC), consisting of the GPCMV gH, gL and GP129, 131 and 133 (homologs of the HCMV PC), enabled generation of a live, attenuated vaccine that was protective against congenital GPCMV transmission, but the vaccine was unacceptably virulent. In parallel studies we observed that viral challenge of pregnant dams with preconception immunity to the prototypical GPCMV strain 22122 with a heterologous `clinical isolate', the CIDMTR strain, resulted in both maternal re-infection and congenital transmission of the new challenge strain. These findings, as well as clinical observations in HCMV-infected women, suggest that a vaccine may need to perform better than `natural immunity' in order to protect against reinfection. We hypothesize that a rationally designed, PC-intact, but disabled infectious single cycle (DISC) vaccine will provide superior protective immunity, including against re-infection, compared to subunit gB and to `natural' immunity. In Aim 1, we test this hypothesis by comparing a PC-positive vaccine, attenuated by deletions of GPCMV-encoded MHC I homologs and a viral protein kinase R (PKR) evasin, versus MF59- adjuvanted recombinant gB. We will examine the magnitude of protection after both primary maternal infection and re-infection of dams with preconception immunity following challenge with a heterologous strain during pregnancy. We will further optimize CMV vaccines in Aim 2 by developing DISC vaccines, using a PC-intact virus with destabilizing domains fused to essential viral proteins. In contrast to HCMV vaccines in clinical trials, we hypothesize that DISC vaccines in which viral replication is blocked downstream of DNA replication will progress to late gene expression and produce a broader array of immunogenic proteins. Aim 3 will evaluate whether deletion of viral antagonists of host cell defenses such as protein kinase R (PKR) can aid in vaccine design. We will test the hypothesis that GPCMV encodes a second PKR antagonist, elimination of which would be expected to generate a safe and effective vaccine. In addition, we will overexpress dsRNA in the vaccine construct with the aim of improving safety and immunogenicity. Vaccines will be evaluated for attenuation and immunogenicity and optimized vaccines will be used to identify immune correlates of protection, including non- neutralizing functions of IgG. Our studies will explore the overarching hypothesis that immunity conferred by a CMV vaccine can be superior to `natural' preconception immunity, a highly relevant issue for HCMV vaccines.