Reconstructing the invasion history of Lyme disease in North Americ

Over the last half-century, previously undescribed tick-borne pathogens including the Lyme disease bacteria, Borrelia burgdorferi, have rapidly spread across the Northeast and Midwest United States. Lyme disease is now the most commonly reported vector-borne disease in North America, with over 300,000 estimated cases each year in the United States. Despite its epidemiological importance, many questions remain about this ongoing invasion. Does the observed spread of human cases reflect the ecological spread of the Lyme disease bacteria or does it reflect changes in case reporting and recognition? How do ticks and tick-borne pathogens spread across space and why are tick-borne pathogens currently invading the US? A better understanding of the ecological and evolutionary history of Lyme disease in North America will inform predictions about its future spread and how control measures might be implemented.;Reconstructing the invasion of Lyme disease is challenging because B. burgdorferi circulates in an enzootic cycle; humans are only incidental hosts. This means that reported cases of disease may not reflect the underlying ecological spread of B. burgdorferi. Pathogen genomes offer an alternative data source for reconstructing the history of pathogen invasion. However, this requires large population-scale samples of pathogen genomes that are difficult to generate from field samples. Further, for pathogen genomes to be informative, pathogens must evolve on similar timescales to ecological spread.;My dissertation work integrates diverse data sources--human case reports and pathogen genomic data--to reconstruct the history of B. burgdorferi in North America. In Chapter One, I present a spatio-temporal model for the spread of human cases of Lyme disease and babesiosis, another tick-borne disease, across New England. Our model uses use the best available longitudinal data--human surveillance data--to model the underlying ecological spread of tick-borne pathogens. Our model predicts that tick- borne diseases spread in a diffusion-like manner, at approximately 10 km per year, with occasional long-distance dispersal, likely due to spread by avian hosts. The remaining studies rely on pathogen genomic data. In Chapter Two, I tackle the methodological challenge of generating genomic data from mixed template samples by developing a method to capture multiple pathogen genomes from individual field-collected tick samples. This approach allowed us to efficiently differentiate between pathogen DNA versus tick and other exogenous DNA, enabling efficient deep sequencing and population genomic study. In Chapter Three, I examined the genomic diversity of B. burgdorferi within individual field-collected ticks. I found that 70% of ticks are infected with multiple strains of the Lyme disease bacteria, indicating that humans may be exposed to and infected with more than one strain of the bacteria from a single tick bite. I also find evidence that the Lyme disease bacteria is evolving in response to the immune defenses of its natural hosts (including rodents and birds). Finally, in Chapter Four, I examined patterns of B. burgdorferi genomic variation across space. I find that B. burgdorferi diversity is ancient and predates not only the reported emergence of Lyme disease in humans over the last ~40 years, but also the last glacial maximum, ~20,000 years ago. Ultimately, population genomic data reveal that the recent emergence of Lyme disease in North America is not driven by a recent introduction or evolution of B. burgdorferi. Instead, the recent epidemic of human Lyme disease is likely driven by environmental and ecological changes that have increased the density of ticks, infected ticks, and/or frequency of human exposures to infected ticks in the past century.