Research

Life expectancy in the United States over time (data for Caucasian women). Will pathogen evolution ultimately reverse these health gains of the past? Data retrieved from the National Vital Statistics Reports.

Over the last century, life expectancy in the United States has nearly doubled.  Almost all of this increase has been attributed to the control of infectious diseases.  Improvements to sanitation and hygiene, the development of antibiotics and vaccines, and the local elimination of key diseases have all contributed to improved human and/or animal health.

But pathogen evolution is threatening to undermine these gains of the past.  New pathogens are emerging from animal reservoirs, preexisting pathogens are evolving to be more virulent, and pathogens that were previously controlled by antibiotic drugs or vaccines are evolving resistance to these interventions.

In the Kennedy lab, we work to develop mechanistic understandings of disease ecology, with the goal of understanding the evolutionary forces that shape pathogen populations.  Ultimately, we would like to develop the knowledge base needed to prevent adverse pathogen evolution.

Linking experimental and quantitative disease ecology

The Kennedy lab primarily works on systems with ecological dynamics that are amenable to experimental manipulation and evolutionary dynamics that can occur over observable timescales. We have therefore been developing a new experimental system, nematodes (Caenorhabditis elegans, C. briggsae) and their recently discovered RNA viruses (Orsay virus, Le Blanc virus, Santeuil virus).  The wealth of resources and knowledge regarding nematode physiology, behavior and genetics makes manipulation of the basic ecology readily achievable, while the fast dynamics and high mutation rate of the viruses allow for evolution over short timescales.  It is therefore a powerful system for studying pathogen evolution.

C. elegans viewed under a dissecting microscope.

We are performing experiments to characterize the basic features of disease dynamics in this system.  In particular, we are estimating virus transmission rate, infection-induced host mortality rate, and virus degradation rate.  We are also confronting models of disease ecology with experimental data to determine whether pathogen transmission is frequency or density dependent. Using this empirical system in combination with mathematical models, collaboration and other data sources, we are exploring questions fundamental to disease management.  In particular, why does evolution often result in pathogen strains that harm their hosts?  Why is a pathogen population able to evolve across a host species boundary in one setting but not another?  Why does pathogen evolution more readily undermine some public health and animal health interventions than others?

What factors drive the evolutionary failure of interventions?

Time to first detection of human pathogens resistant to vaccines and antimicrobial drugs. Viral vaccines are labeled in purple, bacterial vaccines are labeled in green. Global eradication of smallpox (marked as a blue circle), ended the opportunity for resistance to emerge (blue line). The seasonal influenza vaccine is routinely undermined by antigenic evolution, evolution that occurs even in the absence of vaccination (dotted line). Several dates could be debated, but the general pattern is robust: resistance to drugs occurs more readily than resistance to vaccines. Reprinted from (Kennedy & Read 2017).

Evolutionary theory has taught that populations adapt when there is a combination of heritable diversity, selection and time.  Such adaptation can be highly beneficial to the survival of a population after an environmental perturbation, so called evolutionary rescue, but population extinction is often the desired outcome when interventions are applied to infectious disease systems.  Indeed, adaptation can undermine the benefits of public health and animal health interventions through the evolution of resistance.  Some intervention strategies lead to the evolution of resistance faster than others, and it would be useful to understand why.  General answers to this question have thus far been equivocal because the details of systems have often altered the conclusions of individual studies.

We have argued that the comparison between antimicrobial drugs and vaccines might get around these system specific details because with few exceptions pathogens tend to evolve resistance to drugs and tend not to evolve resistance to vaccines.  Several factors may explain why this difference exists (Kennedy & Read 2017, Kennedy & Read 2020).  We combine mathematical models, lab experiments (Honors thesis: Aluquin 2023), and meta-analyses of data from the field (Bhattacharya et al. forthcoming) to quantify the risk factors that determine the rates of intervention failure due to evolution.  Ultimately, our goal is to quantify how different biological details affect the rate of evolutionary failure, so that new intervention strategies can be developed that reduce these rates.

What are the consequences of transmissible vaccines on disease ecology and evolution?

Vaccination can be one of the most efficient and effective tools for controlling the burden of infectious diseases, but in many settings, such as for wildlife diseases or farm animal diseases, logistical and economic hurdles make it impractical to vaccinate large enough fractions of hosts to achieve herd immunity. Transmissible vaccines, defined as vaccines capable of disseminating from vaccinated to non-vaccinated hosts, offer one potential solution to these challenges by amplifying the impact of vaccination campaigns. However, transmissible vaccines are not without risk. Reversion to virulence or recombination with wildtype pathogens could cause transmissible vaccines to make matters worse or complicate elimination efforts. We are quantifying the effects of transmissible vaccines on disease ecology and evolution using the example of an economically important, naturally transmissible vaccine currently in widespread use on poultry farms.

Marek’s disease, a poultry-specific disease that is a threat to sustainable chicken and egg farming, is currently controlled by the “Rispens” vaccine, a live, attenuated vaccine that has been widely used for two decades. Recent experiments have found that this vaccine is capable of efficiently transmitting from vaccinated to non-vaccinated chickens. In addition, advances in whole genome sequencing has revealed the presence of recombination between the vaccine virus and wildtype virus, which is concerning given that the vaccine virus harbors highly virulent forms on the oncogenic meq gene. In collaboration with Venu Nair (Pirbright), Yongxiu Yao (Pirbright), and Moriah Szpara (PSU), we are working to develop general models of transmissible vaccination, experimentally characterize vaccine transmission and its impact on virus transmission, genetically characterize vaccine and virus evolution, and model the overall impact of Rispens vaccination on Marek’s disease virus.

How will the virulence of a disease change over time?

Emerging infectious diseases pose threats to wildlife, agriculture, and human health.  The emergence of an infectious disease is often the result of a host jump, in which a pathogen gains the ability to replicate and transmit in a new host species.  Host jumps are likely to be associated with rapid evolution as the pathogen adapts to the novel host. This evolution has been documented at the genotypic level, but at the phenotypic level, much less is known about the causes and consequences of this adaptation.  A key question is whether pathogens will evolve increased or decreased virulence (host morbidity and mortality) as they adapt to new hosts. The orthodox view is that pathogens evolve decreased virulence as they adapt to their hosts, but in theory, virulence can rise or fall depending on the details of the system. The key knowledge gap is in identifying which details matter.

Raceways of a trout farm in close proximity to a river.

In collaboration with Andrew Wargo (VIMS) and Gael Kurath (USGS), we are working to understand the virulence evolution that occurred in the fish pathogen infectious hematopoietic necrosis virus (IHVN) after it jumped hosts from wild salmon to farmed trout. In this project we take advantage of 1) a library of >1000 viral isolates that span the pre- and post-jump periods, 2) parallel independent trajectories of virulence evolution after two independent host jumps, 3) repeatable field-relevant assays that could be used to quantify virulence and transmission potential in the ancestral and novel hosts, 4) molecular tools for investigating genomic changes associated with virulence and fitness, and 5) a mechanistic understanding of the host-pathogen ecology that could be translated into mathematical models. Using this integrated approach, we are characterizing how virulence changed following the host jumps, and we are elucidating the ecological factors that drove this evolutionary trajectory.

What ecological factors facilitate the emergence of novel pathogens?

The evolution described in the IHNV system above took place after a pathogen host jump, but key ecological and evolutionary details must have aligned to facilitate the initial jump event.  The consequences of host jumps can be dire, but host jumps also tend to be hard to predict, because they are almost always studied retrospectively and sample sizes tend to be extremely small.

The nematode virus system offers the possibility to study host jumps prospectively, under repeatable conditions by experimentally allowing host jumps under controlled ecological settings.  All three viruses are host-species specific, but they have a continuum of infectivity across different worm strains within a species.  By co-rearing worm strains of differing susceptibility and by manipulating and modeling the host-pathogen ecology, we are exploring the frequency of host jumps and the path of subsequent evolution.  By performing these studies under reproducible lab conditions we are able to experimentally challenge our models and theories.  These systems thus offer tremendous power for exploring the ecological and evolution factors underlying host jumps.