Maxwell Farrell (CVR, UoG), Sebastien Calvignac-Spencer (HIOH, HZI), Richard Orton (CVR, UoG), Eva Perez & Simon Gubbins (Pirbright Institute), Roman Biek (SBOHVM, UoG)

In order to sustain transmission, viruses need to constantly establish new infections in susceptible hosts. Which host and viral factors allow a virus to be successfully transmitted, and which factors constrain this process, are fundamental questions that underpin our understanding of viral epidemiology and outbreak control. Increasing evidence further points to some key factors, such as immune priming, to be contingent on the order and timing of previous viral exposure, making this an important area for further work. However, studying these contingencies from observational human data tends to be challenging and resource-intensive. Here, we are using a uniquely rich sample and data set from a closed animal population (African buffalo) where transmission was allowed to occur naturally and where serological and virological sampling was carried out consistently over more than three years. We apply powerful targeted sequencing approaches to characterise viral diversity present within infected individuals and to reconstruct transmission networks of ‘who infected whom’. Combining this information with time-specific information about virus exposure at the individual level, we test how exposure history, contact patterns, and competition among virus strains contribute to transmission success or failure, thus providing insights into the mechanisms of virus persistence at the host population scale. 

Liam Brierley (CVR, UoG), Jan Gogarten (HIOH, HZI), Jordan Bone (CVR, UoG), Lu Lu (UoE)

Traditional epidemiological tracing of the animal host origins of a new zoonotic virus is notoriously challenging. Advances in machine learning now allow algorithms to identify the likely host of a virus directly from genome sequences. Algorithmic host predictions that do not match the current host are often considered to indicate a previous host, though no systematic investigation has explored how reliably these outputs can act as evidence of historical host shifts. 

By combining machine learning with macroevolutionary phylodynamics, we will assess historical host shifts along the influenza phylogeny. Machine learning algorithms will be trained on sequences to classify broad influenza host type (e.g., birds, humans, swine, equine) and phylogenetic host reconstruction carried out in parallel. By extracting the most informative genomic motifs used as features by machine learning algorithms and considering them as quantitative traits along the phylogenetic tree, we will assess convergence between shifts in genome traits and shifts in host type. 

Findings will contribute to phylogenetically-informed guidance of interpretation and follow-up analyses for the rapidly expanding field of computational viral host prediction - an essential step if we are to better understand the fundamental evolutionary drivers of host shifts and potential future zoonotic emergence events. 

Vanessa Herder (CVR, UoG), Ulrich Kalinke (Twincore, HZI), Ana Da Silva Filipe (CVR, UoG), Andreas Pavlou (Twincore, HZI), Inken Waltl (Twincore, HZI)

Infections with many re-emerging and newly emerging viruses can infect the central nervous system (CNS) and may cause viral encephalitis. Often viral encephalitis comes with severe symptoms and many surviving patients develop severe post-encephalitis neurological sequelae. Currently, the therapeutic arsenal to treat neurotropic virus infection of the CNS is very limited.

The nasal epithelium that contains olfactory sensory neurons is constantly exposed to a broad range of viruses that may gain access to the CNS under specific conditions. Upon virus propagation within the olfactory bulb (OB), microglia are activated, which is associated with the induction of morphological changes, proliferation of microglia and recruitment of the cells to sites of infection within the OB. Full microglia activation is needed to restrict virus dissemination within the CNS. Recent single-cell RNA sequencing data from one of our labs suggest that microglia and infiltrating myeloid cells within the OB present a highly diverse transcriptomic profile. Therefore, we plan to address the transcriptomic profile of CNS myeloid cells by spatial transcriptomics. We hope to better understand at which anatomical sites microglia with different activation signatures are located relative to virus-infected neurons in our virus infection model with vesicular stomatitis virus (VSV).

Antonia Ho (CVR, UoG), Berit Lange (HZI), Louisa Pollock (SMDN, UoG), Brian Willett (CVR, UoG)

RSV is the key cause of acute respiratory infection (ARI) in young children worldwide, and a common cause of severe respiratory illness in older adults. New immunisation strategies against RSV promise to greatly reduce the enormous burden RSV places on health services during seasonal peaks. These include two newly licensed recombinant pre-fusion F protein vaccines and a novel monoclonal antibody, nirsevimab. Assessing the impact of these interventions on population RSV dynamics will be crucial. The UK and Germany are proposing different infant immunisation strategies, with both countries planning to vaccinate the elderly, thus providing a unique natural experiment to compare the impact of their respective interventions between similar populations. Robust methods to assess population RSV immunity and transmission dynamics are therefore critical.

This project aims to combine CVR and HZI expertise to characterise the seroprevalence of RSV in a German adult population cohort (MuSPAD) and a Scottish mother-infant cohort (MISC) before the introduction of immunisation strategies. Additionally, we will compare RSV antibody responses between a binding assay with a pseudotyped virus neutralising assay, model age-specific antibody waning, build a common minimum dataset between the cohorts, and develop a protocol for comparative analysis of immunisation effects in these population groups. 

Margaret Hosie (CVR, UoG), Kristin Metzdorf (HZI), Lee Sherry (CVR, UoG), Josef Penninger (HZI)

The process of developing a promising vaccine candidate for use in the clinic can be a long, complex and difficult process. One major hurdle in vaccine development occurs between identifying a vaccine candidate in pre-clinical animal models and testing in human clinical trials, and this is where a significant number of candidates fail. A key factor in these failures is that animals can provide an indication of the immune response to infection, but they cannot replicate the human immune response. Therefore, to better understand the human immune response to infection and vaccination, we propose to establish a 3D cell culture model of human tonsils. This exciting new platform takes healthy human tissue following tonsillectomy and uses it in the lab to reproduce the natural immune response following exposure to a virus or vaccine candidate. In response to the recent increase in measles cases in the UK and worldwide, we propose to test our new model system for its ability to reproduce the immune response that is seen when people are vaccinated using the measles-mumps-rubella (MMR) vaccine. Overall, we believe that this model system has enormous potential to develop improved vaccine candidates. 

Ed Hutchinson (CVR, UoG), Christian Sieben (HZI), Joe Grove (CVR, UoG), Lauren Orr (CVR, UoG)

Binding and entry are the critical first steps in viral infection. Despite the enormous clinical and veterinary impacts of influenza A viruses (IAVs), we do not fully understand the intricate process by which they bind to receptors and induce endocytosis. The Sieben group (HZI) recently used novel imaging methods to show how IAV virions explore the cell surface, bind to receptor clusters, and then locally enrich receptors to trigger endocytosis. As this work used low concentrations of spherical virions, important questions remain: (i) do the filamentous IAV virions characteristic of natural infections enter through the same route, and (ii) are the very high concentrations of virions shed locally during infections endocytosed in the same manner? These interests synergise with those of the Hutchinson group (CVR), who have been developing virological methods to study IAV infectivity. Together, we will answer these questions, using fluorescently-labelled virions to image IAV infections at the single-cell level. Our project combines expertise in molecular virology with advanced fluorescence imaging to study virus-receptor interactions and cell entry in a situation that mimics in vivo infections. As well as resolving ongoing research projects in both groups, it will lay a foundation for further collaborative work on this topic. 

Massimo Palmarini (CVR, UoG), Julia Port (HZI), Arvind Patel (CVR, UoG)

SARS-CoV-2 must be exhaled in respiratory droplets to be transmitted via the airborne route over long distances. Exhalation of SARS-CoV-2 in respiratory droplets is however contingent upon the anatomical distribution and tropism of SARS-CoV-2 within the respiratory tract. It is of critical importance to understand the intricate processes governing virus airborne transmission in order to control COVID-19. We aim to address this fundamental issue by combining our complementary expertise at the CVR and HZI. To this end, we will utilise recombinant viruses in a competition-based in vivo transmission model. This approach, represents an optimal solution that circumvents some of the inherent issues associated with in vitro studies or traditional infection models. Our study will provide data to conclusively determine which section of the respiratory tract needs to be targeted for therapeutics or vaccine-induced immunity. Furthermore, this study will provide an intellectual and experimental framework that will be useful in understanding the transmission of other respiratory pathogens. 

Daniel Streicker (CVR, UoG), Fabian Leendertz (HIOH, HZI), Nardus Mollentze (CVR, UoG), Lorenzo Lagostina (HIOH, HZI), Alice Broos (CVR, UoG), Maxwell Farrell (CVR, UoG)

Human modification of natural environments is increasingly linked to viral zoonotic emergence. The mechanisms underpinning these emergence events include the creation of novel human-animal interfaces, ecological changes that alter the incidence of infection in animal reservoirs, and immunological changes within animal reservoirs which influence the intensity of viral shedding. This pump priming project explores how human actions alter viral dynamics in animal reservoirs by applying a massively multiplexed serological assay designed by CVR investigators to wildlife samples collected by HZI and CVR investigators. Using a land use gradient in Côte d’Ivoire, we will test how transitioning from pristine forests, to agricultural fields, to villages alters viral exposure histories in bats and rodents. We hypothesize that ecological homogenization will reduce viral antibody diversity but will increase the seroprevalence of remaining viruses due to reduced host condition or behavioural/population dynamic changes in altered animal communities. Next, by adding bat samples from a variety of environmental conditions in Peru, we will test the hypothesis that virus infection varies by species in ways that exacerbate risk under anthropogenic conditions. These studies and associated researcher exchanges will catalyze further CVRHZI collaborations and show how cutting-edge techniques in virology can demystify on-the-ground zoonotic risk.