A Systematic Investigation into Human Cytomegalovirus Gene Function

Worldwide, most people are infected with cytomegalovirus (CMV) without realising it. CMV is closely related to herpes simplex, the virus that causes cold sore. Like herpes simplex, CMV infections are for life and must be constantly controlled by our immune system. Disease occurs when this control breaks down.
CMV disease is a major problem in hospitals; primarily due to its capacity to cause life-threatening infections in immunosuppressed patients, particularly bone marrow, kidney and heart transplant recipients plus patients with HIV-AIDS. CMV disease can be manifest in virtually any organs (e.g. lungs, intestine, lung, brain, eye, ear). Moreover, researchers have recently become extremely concerned that low grade CMV infections are commonly associated with the most common and deadly brain tumour (Glioblastoma multiforme). It has been hypothesised that inflammation associated with the CMV infection could be driving such tumours. Critically, CMV is able to cross the placenta to infect the foetus, and does so in ~1% of pregnancies. Congenital infection can induce miscarriage, is a common cause of deafness and in its more severe form causes brain damage. In brutal terms, the cost in patient care alone is $1.86 billion per year in the US. Congenital CMV disease is driving vaccine development. More subtly, CMV infection alter the make up of immune cells of apparently healthy carriers and the implications of this are not known. CMV is a major human pathogen that needs to be researched and merits a much higher public profile.

CMV is the most complex human virus containing nearly 200 genes, of which >124 genes are not required to replicate the virus. Research by us, and others, show many are controlling our immune system. As most of us carry this pathogen throughout life, it is important to find out what these genes are doing. To address this issue we first defined the genome of the clinical agent and then developed systems to screen for CMV gene function: we generated i) a vector library capable of expressing all CMV genes in cells individually and ii) a panel of CMV viruses deleted in large blocks of genes. Combining these resources has enabled us to rapidly map a substantial number of CMV immune evasion functions. However, we would also wanted to know how they worked. Collaborations have been established with researchers in Harvard and Cambridge to exploit state-of the-art proteomic systems to track expression of >700 cell surface and >7000 intracellular proteins during virus infections. These technologies combine to reveal exactly how individual CMV immune evasion genes act.

Natural Killer (NK) cells are white blood cells critical in controlling CMV disease. We are currently systematically screening all CMV genes for their capacity to modulate NK cell function. We seek funding to first analyse the mechanism of action of 7 novel NK evasion genes (RL10, UL4, UL20, UL24, UL26. US18 and US20). The CMV genome is organised into 'families' of related genes that have arisen by duplication of a common ancient precursor. The US12 gene family has 10 members arranged sequentially on the genome and is uncharacterised. We have mapped two distinct NK cell evasion functions to the US12 family members US18 and US20. US18 target multiple cellular proteins; some co-operatively others independently. We seek to utilise proteomic technology to systematically investigate all 10 members of the US12 gene family to determine their function and the extent of their interdependence. An 'accordion gene expansion' occurs when a gene is amplified to enhance its function to counter selective pressure. We suspect the US12 family of acting as an 'accordion gene expansion'. We also propose to similarly follow a second large gene family (RL11) strongly implicated in immune modulation, but where as yet there is no evidence of co-operation. These studies will provide major insights into CMV pathogenesis and evolution.

We defined the genetic content of a circulating HCMV strain (Merlin) and generated an infectious BAC clone to provide a secure basis for studies into HCMV pathogenesis. The infectious HCMV Merlin BAC clone provides a reliable source of wild-type genes and genetically intact, defined clonal virus. Detailed transcriptional analyses underpin the definition of HCMV canonical genes. To permit functional screening, 170 HCMV protein-coding genes were inserted into an adenovirus vector and a panel of HCMV 'block' deletion mutants spanning >50% of genome assembled. Ongoing screening of HCMV genes for recognised and novel NK cell modulatory functions will be completed prior to initiation of the project. The literature details 7 HCMV NK evasion functions; to date, we have identified 8 more we wish to characterise.

NK cells recognise a combination of activating and inhibitory ligands on the surface of targets cells. We have therefore developed Plasma Membrane Profiling (PMP) to simultaneously track the expression of >700 cell surface viral and cellular proteins in multiple samples. Using bespoke HCMV mutants, PMP will be used to characterise the mechanism-of-action of novel NK evasion functions in the context of virus infection. Where necessary mechanism will be further explored in whole cell proteomic analyses and SILAC-immunoprecipitation. The contribution of host proteins will be elucidated by genetrap retrovirus-mediated mutagenesis in a haploid cell line (KBM7).

Remarkably, the HCMV genome contains multiple sets of homologous genes that have been assigned to gene 'families'. Functional screening with HCMV 'block' deletion mutants highlighted roles for multiple members of the RL11 and US12 gene families in immune modulations. Using genetic screening and proteomic technologies pioneered to characterise NK modulatory functions, we shall seek to systematically investigate conservation and divergence of gene function across the two largest HCMV gene families.

This is a fundamental research project aimed at elucidating basic mechanisms of virus pathogenesis rather than targeting an immediate translational endpoint. Nevertheless there is enormous public interest in Support Groups of parents whose children are affected by the consequences of congenital HCMV infections. Members of such groups regularly attend and participate at primary research meetings and are extremely interested in all research on this virus. Parents tend to be frustrated that HCMV is clearly a very important pathogen yet it is perceived to be 'obscure'. Great solace is taken by these groups in all progress made by HCMV researchers.
HCMV was identified as a leading vaccine candidate by the US Institute of Health. The content of HCMV virions has been ascertained only for one laboratory-adapted strain, and is known to be incomplete. For example RL13, a member of the RL11 family (to be studied in this proposal), is mutated in effectively all passaged HCMV isolates yet encodes a hypervariable virion glycoprotein that potentially contains neutralising epitopes. Most of the genes studied in this application are uncharacterised glycoproteins and although few are likely to be targets for neutralising antibody, a number may be targeted by ADCC. The results of this study may ultimately be important to commercial enterprises interested in HCMV vaccine development.
HCMV has been mooted as a potential vaccine carrier, most notably for HIV, TB and various tumour antigens. The strongest T cell response made to any infectious agent is made to HCMV. The Rhesus CMV (RhCMV) model has demonstrated the capacity of CMV to superinfect seropositive animals and in this environment induce strong T cell responses to expressed antigens. The RhCMV vaccine vector is unusual in being capable of inducing an effector:memory response that can not only elicits protection against and SIV challenge but will persist/inflate over time. Studies using RhCMV mutants have shown that the quality of the T cell response induced is extremely sensitive to both the tropism of the carrier virus and expression of immune evasion functions. The RhCMV vector is a most encouraging model for HCMV. In developing HCMV as a vaccine carrier there is a clear need to modify its genome to render it non-pathogenic while still maintaining suitable delivery and expression of antigen. Information on HCMV gene usage clearly has extreme relevance to experiments aimed at trialling RhCMV in vaccine development. We have observed conservation of function between NK evasion function UL141 in HCMV and its homologue in RhCMV (Aicheler, unpublished); whereas the MHC-I homologues (UL18 and UL142) are missing from RhCMV. Interestingly, the RL11 and US12 gene families are found in RhCMV, including direct homologues of US18 and US20 (Lesniewski et al, Virology 2006: 354, 286). Functional information on these two large gene families will be invaluable for HCMV/RhCMV vaccine vector development.

While Ganciclovir is widely used in the treatment of HCMV disease, the drug has recognised cytotoxicity and resistance emerges rapidly. All information on virus gene function has the potential to inform medicinal chemistry. This project has the potential to provide insight into the function of some 28 HCMV genes. Of particular interest are the ten members of the US12 gene family that encodes a group of 7 transmembrane spanning receptors, a large superfamily that include G protein coupled receptors (GPCRs). Over 30% of marketed drugs target GPCRs.