Systematic characterisation of the HCMV pathogenicity domain spontaneously lost from commonly used passaged strains

Human cytomegalovirus (HCMV) is a herpes virus, related to the virus that causes cold sores. HCMV infects almost everybody, without causing obvious disease and is usually mistaken for a mild flu. Most people never realise they are infected, but in some very specific cases, they will experience serious illness. HCMV is one of the few viruses that can cross placenta and infect unborn babies. Every year in UK around 5000 babies are infected this way. Of these, around 50 die and 1000 develop disabilities like hearing loss, blindness or brain damage. In fact, HCMV is responsible for most cases of developmental disorders in babies. In adults, HCMV is a serious problem for patients with HIV-AIDS and transplant patients. Even with care, a high proportion of patients will experience widespread disease, leading to damage to the lungs, eyes, gut or transplanted organs, and severe complications or even death. HCMV screening, antiviral and supportive treatments are top priorities for clinicians managing all these patient groups. It is estimated that caring for HCMV-patients in UK costs around £500 million every year.
So far, scientists haven't developed a vaccine or drugs that cure HCMV or treat it without severe side effects. Perhaps since we don't know enough about HCMV. It is known that all cases of HCMV disease have one thing in common - an immune system that does not work at its best. HCMV infection lasts for life and our immune system keeps it constantly in check to prevent disease. But the virus always manages to hang around and takes advantage of any opportunity when the immune system breaks down, to grow and cause illness. Even in apparently healthy infected people, the virus changes their immune system with unknown consequences for our long-term wellbeing. Studying HCMV's interactions with our immune system is therefore very interesting and important. So far, we have been quite successful and discovered how 10 HCMV genes stop white blood cells called Natural Killer cells from killing the infection.
This search is hard, because HCMV is the largest and most complex human virus. HCMV keeps over 100 genes just to interfere with our immune system and each gene has evolved many functions. As these functions are only needed in people they are often 'lost' from the virus genome when grown in cells, for example the large gene region called UL/b' which we wish to study. We don't yet understand how or why HCMV loses its genes in culture, but it's quite fortunate for us. For instance, one of the first genes lost when the virus is grown in cells, called UL141, is the most potent immune inhibitor. Focusing on genes that 'get lost first' is a very effective way of deciding which ones to study. UL141 is in the centre of the 22 genes that make up the UL/b' region. We have been interested in studying this region for a while and made a lot of effort developing technology for it. First, we 'decoded' or sequenced the genetic material of virus from patients and followed how it changed during culture. We found that UL141 gets lost first and is then followed by the rest of UL/b'. To make sure that the virus for our experiments doesn't change, we keep it preserved in bacteria and only grow it once in tissue culture when we need a stock of virus. Because the virus genome in bacteria is easy to cut and paste, we were able to make a library of 22 virus mutants, each missing one UL/b' gene. This way, if we mark each virus sample differently, we can analyse them all together using an instrument called a mass spectrometer. This is a good way to look at all proteins in the cell at the same time and can tell us what our virus proteins do to the cell and how they do it. We therefore always start with proteomics when looking at functions that we know nothing about.
Ultimately, our studies will tell us how the virus works, how it interacts with our immune system, what is the best way to make a vaccine, or what is the best way to kill the virus.

Prior to this application, the genome and transcriptome of intact human cytomegalovirus (HCMV) strain Merlin was first defined (with Dr. Davison, Glasgow). Bacterial artificial chromosome (BAC)-cloned Merlin DNA provides a stable source of infectious virus. For functional screens, 10 HCMV 'block' mutants with deletions of 2-9 genes and adenovirus (Ad) vector library expressing all HCMV ORFs were constructed. Whilst the Ad library is routinely used to screen individual HCMV functions, single HCMV gene deletion mutants are used as gold-standard reagents. All 22 single deletion HCMV UL/b' mutants (UL131A-UL150A) have been generated, validated by whole genome sequencing (Dr. Davison), and functionally, leading to discovery of UL148, as described in this application. These reagents allow for effective loss of function screens of the entire UL/b' region.
Mass spectrometry (MS) technologies were developed specifically for HCMV proteomics (with Dr. Weekes and Prof. Lehner, Cambridge), for analysis of plasma membrane (PM) and whole cell lysate (WCL) proteomes over time, deletion mutant analysis is a rational next step. Proteomics employs 10-plex Tandem Mass Tag (TMT) labels for higher throughput analysis. Proteomics is now established in the lab and used to analyse 10 HCMV US12 family deletion mutants confirming its feasibility. Dr. Fielding (named RA) is trained in proteomic techniques and analysis. Access to HpRP-HPLC, Orbitrap Fusion MS etc. is via collaboration with Dr. Weekes and Prof. Lehner (letter of support attached).
NK cell techniques are well established, including NK expansion/cloning, multicolour phenotyping, matched effector/target cells. Delineating NK function is challenging, due to multiple activating and inhibiting receptor inputs. For the first time, we are able to examine NK functional effect against changes in PM profile of 1,200 cell surface markers and connect functional screens and mechanistic studies.

Direct and immediate beneficiaries from this research will be the scientific community. Interest in HCMV proteomics databases for instance is overwhelming. Very few laboratories have the capacity to perform complex proteomics, however our processed data are freely available for mining and also serve as reference resource for general virology and immunology communities, informing on regulation of thousands of cellular and viral markers. Both on-going and future bioinformatics analyses of these datasets will be undoubtedly hypothesis generating and will inform on mechanisms governing fundamental cellular and immune processes. It is our intention to distribute all future proteomics databases and virus banks freely and as widely as possible. The future databases will also include interactome libraries, extending their utility in HCMV biology. HCMV Merlin is now official NCBI reference strain (http://goo.gl/bCOJah) and has been adopted by ATCC (http://goo.gl/A9nEqb). HCMV has poorly defined clinical symptoms, diagnosis is only possible by laboratory testing. Our HCMV Merlin strain has been adopted by WHO as a diagnostic standard (http://goo.gl/WPMdtH) and is now used in 43 countries to calibrate diagnostic kits. A testing kit using the Merlin standard has also been approved by FDA. Adoption of Merlin standard improved consistency for diagnosis, impacting on screening and management of HCMV disease. HCMV has been identified as a leading vaccine candidate by the US Institute of Medicine. The complete virion content for intact HCMV is however not known, our identification of RL13 and UL141 (unpublished) as virion glycoproteins and their impact will be considerable as both virus function as tropism determinants and immune targets. Clinical trials with HCMV AD169 and Towne strains (lacking UL/b' functions) that showed a striking and complete lack of pathogenicity were validated by recent rhesus CMV (RhCMV) studies, concluding that UL/b' functions are essential for optimal primary infection, replication and horizontal transmission. Because of its unparalleled ability to stimulate potent and persistent effector memory T-cell responses, HCMV is also a strong candidate for vaccine carrier, most notably for HIV, TB and tumour antigens. Recent RhCMV studies have shown that the quality of these T-cell responses are affected by UL/b' functions. HCMV vaccine based on defined manipulation of UL/b' is currently being pursued (Dr. Tong-Ming Fu, Merck, personal communication). To inform on rational design of future vaccine candidates, basic understanding of UL/b' region, proposed in this application, is essential. There is also an enormous public interest in HCMV research in support groups of parents whose children are affected by consequences of congenital HCMV infections. These groups regularly attend and participate at primary research meetings and show considerable interest in all aspects of HCMV research. Parents tend to be frustrated that HCMV is clearly an important pathogen, yet it is perceived to be obscure. Great solace is take by these groups in all progress made towards both understanding of HCMV biology and development of specific therapies.