Evolutionary epidemiology of multi-transmission pathogens in multi-host networksOverview. This project brings together two topics of broad societal and scientific interest: the ecology of emerging infectious diseases and the dynamics of biological networks. The emergence of novel infectious diseases is arguably the most complex ecological-anthropological research frontier in modern biology. With this squarely in view1, we propose to develop a computational model of pathogen transmission and evolution embedded within a multi-scale ensemble of heterogeneous networks, calibrated with a unique data set on avian influenza viruses (AIV) in North America. It is well known that the majority of emerging infectious diseases are zoonotic (diseases of animals) that either spillover into the human population or evolve the capacity for human-to-human transmission, an evolutionary step known as "host shift"2. The AIV system is an ideal model for the kinds of ecological networks we believe underlie the emergence of many human diseases. It is also a crucially important system for another reason: this is the system from which pandemic H5N1 would emerge, should it shift hosts from the wild bird reservoirs in which it currently persists, to infect and transmit between humans.
Influenza: Importance to Society. The importance of understanding influenza ecology to human societies is inestimable. Global mortality of the 1918 pandemic is currently estimated at 50 million to 100 million people or 2.5% to 5.0% of the world population3,4. If a new pandemic achieved similar infection and mortality rates, as many as 330 million people would succumb (based on an estimate of current global population size of 6.65 billion5). While modern vaccines might prevent many infections, the presumed cause of severe disease (and ultimately mortality) in the 1918 pandemic is the development of an unregulated immune response known as a "cytokine storm" that results from positive feedback between cytokines (cell signaling compounds) and immune cells6. Importantly, treatment to control hypercytokinemia is still experimental. As of March 5, 2008, the human mortality rate of spillover H5N1 is about 63% (http://www.who.int/csr/disease/avian_influenza/country/en/). Thus, failure to be vigilant is a recklessness that should not be tolerated by society, government, or researchers, as demonstrated by the high priority placed on pandemic preparedness and the proposed US Pandemic Preparedness and Response Act of 2005.
Of course, clinical treatment and pandemic preparedness are the purview of medical research. What medicine cannot address is the ecological backdrop against which a host shift from birds to humans must occur for an influenza pandemic in the first place2. The proposed research aims to fill this gap. Broadly speaking, the most important outstanding questions are (i) why do host shifts occur when they do? and (ii) how is the virulence of the emerged strain evolutionarily determined? In this proposal, we develop some new hypotheses. Specifically, we think that alternative modes of transmission modify the evolutionary context in which epidemiologically relevant biological traits such as replication, environmental tolerance and infectiousness are subject to selection. We postulate that even small amounts of non-direct transmission might give rise to complicated ecological-evolutionary feedbacks. These, in turn, give rise to greater genetic, antigenic and immunological diversity creating a larger repertoire from which emerging viruses can evolve. If our idea is correct, then understanding the ecological-evolutionary dynamics of low-pathogenic influenza in its natural hosts (waterbirds) is crucial for understanding the evolution of high-pathogenic influenza in humans.
Avian Influenza Viruses. AIV are an ideal system for addressing the questions outlined above because of their ecological complexity: (i) in North America, AIVs naturally infect 90 species of birds from 13 orders, mostly of Anseriformes (ducks, geese, and swans) and Charadriiformes (gulls, terns, and shorebirds)7,8, (ii) their gene pool in aquatic birds provides the biological variation required for the emergence of novel phenotypes and virulence evolution7,9 and, (iii) a seasonally-driven migratory biology. Transmission and maintenance of AIV in wild bird populations is by fecal/oral route in contaminated water10. Replication occurs primarily in the intestinal tract with high concentrations of infectious virus shed in feces. AIVs have been isolated from freshly deposited fecal material, and from unconcentrated lake water9. Thus, the understanding of this disease system is thoroughly grounded in natural history and pathology, and the development of its ecological and evolutionary theory is timely.
Importance to Ecology, Complex Systems, and Theoretical Biology. In addition to the practical benefit of a better understanding of flu dynamics, the proposed project will make intellectual contributions to ecology. Three research frontiers in ecology include:
- Population dynamics in loosely coupled networks of interacting species
- The effect of evolution on population dynamics when evolutionary and population growth processes occur at similar timescales
- The role of individual physiological diversity in population dynamics.
The avian influenza system we are studying addresses each of these questions. Concerning (1), as a multi-host-pathogen system, avian influenza is a special case of the larger question of population dynamics in loosely coupled networks (loosely coupled because there are some strongly interacting species, but most species are relatively weakly interacting). As a special case, it also affords some unique contrasts to the aquatic food webs, plant-pollinator systems, and plant-herbivore systems commonly used for this line of research. Concerning (2), influenza, a single-stranded RNA virus, is enabled by high mutation rates to evolve rapidly. Selection is similarly strong and efficient with a high fitness differential, partly due to the selective pressure of learned immune response in spatially heterogeneous host populations. Hosts, in turn, reproduce at age one year, typically live 3 to 7 years, and reach maximum lifespan of 10 to 20 years under ideal conditions. Thus, all the necessary ingredients are present for ecology and evolution to occur on the same time scale. Finally, the effect of learned immune response means that there is variation among individuals, both by cohort and spatial stratification, in their response to infection and ultimately transmission.
Our proposed program of research will provide a much-needed link between burgeoning body of theory concerning biological networks and a empirically-derived system of significant public health and scientific interest. Further, we believe our conceptual findings on these multi-scale hierarchies will extend to other biological networks, including networks of genes, neurons, social groups, protein-protein interactions and biomolecules1112.
