21-EEID US-UK Collab: Multi-scale infection dynamics from cells to landscapes: foot-and-mouth disease viruses in African buffalo

Lead Research Organisation: The Pirbright Institute
Department Name: Transmission Biology

Abstract

Infectious disease dynamics necessarily operate across biological scales: pathogens replicate within host cells and tissues; they transmit among hosts, and across host populations. As such, functional changes in host-pathogen interactions, such as those affecting pathogen vital rates or host immune responses, can generate cascading effects from molecular to landscape scales. Similarly, variation in pathogen success at larger scales generates selective pressures that feed back to shape pathogen population genetics. Linking pathogen dynamics across biological scales is thus critical to understanding evolutionary trajectories of host-pathogen systems and represents a central challenge in disease ecology.

Multi-scale models of infectious disease dynamics seek to address this challenge by linking mechanistic models representing host-pathogen interactions from cellular to population scales. Developing the mathematical tools for connecting dynamic processes operating at vastly different temporal and spatial scales has been an active focus in infectious disease modelling over recent years. These theoretical innovations have so far not been matched by empirical data generation, providing integrated data streams documenting infection processes in the same host-pathogen system, collected consistently across organizational scales. This mismatch has limited the iterative interplay between theory and data. As such, the full potential of multi-scale disease models has not been realized. If successful, these approaches could provide tools for predicting the spread and persistence of new pathogen variants in natural host populations from variant genetic or phenotypic traits - a question of immediate urgency.

In this project we will investigate viral dynamics from genomic to landscape scales using a suite of foot-and-mouth disease viruses (FMDVs) in their reservoir host, African buffalo, as a model system. FMDVs are highly contagious viruses that cause clinical disease in domestic ungulates, while endemic infections in their wildlife reservoir tend to be subclinical. FMDVs present an excellent model system for merging theory and data on multi-scale disease processes: (i) they are ubiquitous in a common wild host; (ii) viral diversity is high, with three distinct serotypes circulating in the buffalo population essentially independently, and well-differentiated lineages documented within each serotype; (iii) mutations accrue rapidly, providing high resolution for molecular tracing and phylodynamic analysis; and (iv) due to FMDV's importance as a livestock pathogen, the system is tractable with well established methods for virus culture, experimental challenges, diagnostics and quantifying immune responses.

Building on our previous work in South Africa's Kruger National Park (KNP), we will create a data-driven mathematical framework linking viral dynamics across organizational scales, to test if and how dynamics within hosts, at population and landscape scales can be predicted from phenotypic variation among viral lineages. We will investigate viral dynamics at cellular, within-host, population and landscape scales using mathematical models informed by experiments, observational field studies and phylodynamic analysis. In our models, we will establish explicit linkages across scales by allowing dynamics at the smaller scales to determine parameters for infection processes at larger scales.

Technical Summary

In this project we will investigate the dynamics of foot-and-mouth disease viruses (FMDVs) from genomic to landscape scales in their reservoir host, African buffalo. We will develop a suite of mathematical models to describe viral dynamics at different scales (cell-culture; within-host; between-host; and landscape), which will be parameterised and tested using empirical data collected at each scale. Models will be linked across scales by allowing dynamics at the smaller scales to determine parameters for infection processes at larger scales. The limitations of this modular approach will be assessed through identifiability analysis. More specifically, historic and current isolates collected from the buffalo in the Kruger National Park (KNP) in South Africa will be sequenced and a selection of strains will be characterised in vitro. A mathematical model of dynamics in cell culture will be used to assess differences in growth rates among strains. Buffalo will be challenged with a subset of strains and the dynamics of FMDV and immune responses in each animal quantified using a mathematical model. Relationships between the within-host parameters and in vitro traits will be assessed. The challenge experiments will be used to estimate transmission rates for the strains using transmission models and trees inferred from sequence data. These will be incorporated in a within-herd model to predict variation in transmission dynamics amongst the strains and the predictions tested using longitudinal data collected in wild buffalo herds. Data on circulating viral strains and contact patterns among herds will be used to develop a model to predict viral spread in the landscape. Phylodynamic analysis will be used to test predictions of how life history variation among strains defines patterns of viral spread and persistence in the KNP ecosystem.

Publications

10 25 50