Funded Grants

The evolutionary and epidemiological potential of pathogens

‘I simply wish that, in a matter which so closely concerns the wellbeing of the human race, no decision shall be made without all the knowledge which a little analysis and calculation can provide.’ — Daniel Bernoulli, 1760

This wish, made by one of the first mathematical epidemiologists, is coming true. Math has become an invaluable tool in the fight against infectious diseases. It allows public health officials to conduct virtual experiments that would be practically unfeasible or unethical. Controlled experiments, in which one population receives an intervention and the other does not, are impossible as we clearly cannot intentionally introduce disease into populations or withhold potentially lifesaving efforts for the sake of scientific study. Mathematical models provide a way around this difficulty.

In the 18th century, Daniel Bernoulli made simple calculations that precipitated the eradication of smallpox from Europe and William Farr used statistics to discover that cholera spread via water rather than air, leading to more effective control measures. In the 20th century, mathematical epidemiology exploded and contributed to the prediction and control of almost every disease plaguing humans and their agriculture.

While we can effectively prevent or control many of the diseases we know well, we struggle when new ones appear or old ones evolve. Animal pathogens (zoonotics) regularly jump into human populations. In 1918, a new and highly virulent strain of avian flu notoriously moved into humans, killing over 20 million people. Later in the 20th century, HIV crossed from chimpanzees into humans, and has since claimed over 25 million human lives. In the last five years alone, we have caught SARS from civet cats, monkeypox from Gambian rats, Ebola from fruit bats, West Nile from mosquitos that fed on infected birds, hantavirus from deer mice, and possibly variant Creutzfeldt-Jakob Disease from eating cattle with mad cow disease.

The diverse evolutionary patterns of pathogens are puzzling. Why do some zoonotic diseases adapt to humans while others do not? Why does HIV evolve too quickly for us to develop an effective vaccine whereas flu evolves moderately, requiring the annual redevelopment of new vaccines, and measles is so stagnant that we are still protected by a vaccine developed in the 1960’s? How does our own behavior influence these evolutionary patterns? My work is at the intersection of evolutionary biology and epidemiology. I use mathematics to study the complex interplay between mutation, natural selection and disease transmission that leads to the rapid evolution of pathogens.

This project addresses two important factors that contribute to ability of a pathogen to evolve: the mutations that fuel its evolution and the host conditions that force it to evolve.

(1) The fundamental nature of mutations: Mutations can mean the proliferation or demise of a virus. From the perspective of the virus, good mutations can be the key to avoiding recognition by the host immune system or evading therapeutic intervention, while bad ones can be fatal. From our perspective, forecasting such mutations can be the key to developing effective controls—both vaccines that can withstand pathogen evolution and antiviral therapies that, by increasing the frequency of mutations, cause a mutational meltdown of the virus.

Although evolutionary theory says much about the fate of a mutation once it arises, it says nothing about the spectrum of mutations that will arise in the first place. Some mutations will make the organism more fit, others will harm or even kill the organism, and others still will have little if no effect on the essential functions of the organism. One can think of genes as being points in a vast network of all possible genes, connected together by mutations that convert one gene to another. Organisms evolve by moving along the edges of this network. Random mutations push out in all directions, while natural selection confines the population to edges that maintain or increase fitness.

To better understand the spectrum of mutations that drive evolution and the puzzling differences in the evolutionary propensities of viruses, we will build biologically realistic models of the mutational networks that underlie the evolution of ribonucleic acid (RNA) molecules. RNA is an essential molecule, which plays a critical role in protein synthesis and serves as the genetic material for RNA viruses like HIV, influenza and measles. By analyzing these networks and developing statistical methods for inferring the structures of more complex mutational networks, we will gain basic insight into the evolutionary patterns of viruses. This will allow us to explain why some viruses can evolve more quickly than others, to forecast mutations that might arise in the future, and to develop more robust vaccines and therapies.

(2) The host-to-host transmission of mutant pathogens: We will also study how our own behavior—our contact patterns and public health interventions—constrains the evolution of pathogens. Contact patterns can vary significantly from person to person. Nurses and doctors have much higher numbers of close physical contacts each day than do infants and retirees. The number and nature of interpersonal interactions determine one’s risk of catching a disease and one’s likelihood of further spreading the disease. Yet the most widely used mathematical models for predicting the spread of disease largely ignore diversity in contact patterns, and thus are unable to provide satisfactory answers to some important public health problems.

Over the last five years, I have been working with an international team of epidemiologists, public health officials, and physicists to develop powerful methods for predicting and controlling the spread of infectious diseases. The first step in this approach is to build a model of the network of human contacts that can lead to disease transmission. For example, we have modeled contact networks within schools, prisons, hospitals, cities and the entire US. The second step is to use mathematical methods to predict the spread of disease through the community and to evaluate the efficacy of various forms of intervention. Our recent research has suggested more effective measures for controlling pneumonia in health care settings, containing outbreaks of respiratory-borne infections in cities, vaccinating populations against influenza, and preventing long-range transmission of diseases via air travel.

We will extend this approach to study the complex relationship between contact network structure and pathogen evolution. Consider influenza. The spread of flu hinges on the structure of the underlying contact network. After infection, an individual will initially be immune to re-infection. The contact network will therefore thin out over the course of a flu season, as infections effectively remove individuals from future transmission. This change will impede future outbreaks. The flu virus, however, is continually evolving through a process called antigenic drift. In any given year, the circulating strains may be immunologically distinct from past strains, and can thus re-infect some or all individuals that were previously infected. The fate of a new strain will depend on the residual structure of the host contact network, that is, the contacts patterns between individuals who remain partially or completely susceptible to the new strain. Using realistic contact network models for hospitals and cities, we will study the impacts of host contact patterns on pathogen evolution and vice versa. This will enable us to predict the spread of mutant strains, develop vaccination strategies that complement disease-induced immunity already present in the population, and modify our behavior to discourage the evolution of harmful strains.

These investigations are part of our broader research program to study the complex mutational processes and host demographics that fundamentally steer pathogen evolution. This grant is providing an exciting opportunity to fit together pieces in the puzzle of pathogen evolution, and take steps towards more effective prediction and control of rapidly evolving diseases.