Human cytomegalovirus (HCMV) is a ubiquitous herpesvirus which infects 50-100% of humans worldwide. HCMV causes a lifelong subclinical infection in immunocompetent individuals, but is a serious cause of mortality and morbidity in the immunocompromised and in neonates. Like other herpesviruses, HCMV establishes latency in specific cell types following primary infection, and reactivates periodically during the lifetime of the host. One important site of HCMV latency is early myeloid lineage cells, including hematopoietic progenitor cells and monocytes, in which the critical viral lytic promoter, the major immediate early promoter (MIEP), is repressed. This is mediated by a combination of host and viral factors, including the viral G-protein coupled receptor US28.
Here, I explore mechanisms by which US28 optimises host cells for latent carriage. Using an unbiased proteomic screen, I have assessed changes in total host proteins induced by US28 and find that interferon-inducible genes are downregulated by US28. I validate that MHC Class II and two PYHIN proteins, MNDA and IFI16, are downregulated during experimental latency in primary human CD14+ monocytes. By overexpressing IFI16, I show that IFI16 can activate the viral major immediate early promoter and immediate early gene expression during latency via NF-κB, a function which explains why downregulation of IFI16 during latency is advantageous for the virus. I also show that MNDA is a potential restriction factor for HCMV latency. Since PYHIN proteins are sensors of double stranded DNA, I also investigate whether US28 interferes with the sensing of dsDNA. I also examine the antiviral potential of two US28-targeting reagents during HCMV latency. Lowering latent viral loads in solid organ or hematopoietic stem cell transplant donors and recipients is likely to lead to lower incidence of HCMV-disease in transplant patients. The first reagent, a US28-specific nanobody, inhibits US28 function and partially reverses latency. This leads to lytic gene expression, and subsequent recognition and killing of latently infected cells by naturally existing T cells from seropositive individuals. The second, a US28-specific immunotoxin, has previously been shown to directly kill latently infected cells. I show that new derivatives of this immunotoxin are more efficacious and can kill latently infected cells after a short incubation, paving the way for their use in ex vivo normothermic perfusion of solid organs from seropositive individuals prior to transplantation.