Broadly speaking, we are interested in understanding the ecological and evolutionary processes that govern pathogen severity, and the implementation of this information to improve disease management. Below are some of the general research questions we are addressing, followed by detailed descriptions of the research topics for the avid viewer.
Whether pathogens will evolve a virulent or benign state towards their host has long been debated among the scientific community. Before entering this debate, it is important to understand the subtleties of pathogen fitness. Like all organisms, pathogen fitness is determined by how much genetic material is passed onto the next generation. However, since pathogens depend on their host for survival, they must transmit to a new host before their host dies, or else their genetics will die out with their host. As such, pathogen strains with the greatest transmission success will have the highest fitness. Conventional wisdom ascertains that virulence reduces the transmission duration of a pathogen by decreasing the host lifespan, and thus hypothesizes that pathogens should evolve towards low virulence, so as to maximize transmission. In contrast, there is an alternative hypothesis that virulence is actually a consequence of pathogen replication, such that pathogens which reach higher densities in their host will cause greater levels of virulence. If pathogen replication rates are associated with transmission rates, virulence provides pathogens with a transmission advantage and the evolution of increased virulence is predicted. Currently, the most widely accepted paradigm is that virulence provides both the fitness cost of reduced host lifespan and benefits such as increased replication. Therefore, intermediate levels of virulence are predicted which balance this cost and benefit to maximize transmission. This is often referred to as the virulence trade-off hypothesis. Under the virulence trade-off hypothesis, the intermediate level of virulence which evolves will depend on the system and ecological conditions (Fig. 1). A great deal of our work is focused on determining the positive and negative associations between pathogen fitness and virulence in the systems we work on. We quantify how a variety of pathogen traits such as host entry, replication, competitive ability, and shedding influence overall transmission, and characterize what association these individual traits have with virulence. This work is aimed at determining what ecological conditions might select for increased virulence evolution, so as to bolster management strategies for preventing it from occurring.
Farming and Virulence Evolution
The ecology of pathogens is very different in farmed host populations compared to wild populations. Farmed hosts tend to have altered densities, reproduction cycles, cohort structure, genetics (through inbreeding and or selective breeding), and disease management than their wild counterparts. It is likely that these and other farming practices could have significant impacts on the evolution of pathogen virulence. For example, increased host density in farming is hypothesized to select for the evolution of increased virulence. The logic for this hypothesis stems from the virulence trade-off hypothesis. Pathogens must moderate their virulence so they don’t kill their host before they transmit. However, when hosts are crowded, transmission opportunities are high, and the cost of virulence is low. Therefore pathogens will maximize their overall transmission by increasing replication, and thus virulence. We are exploring how these and other farming practices such as culling might shape the evolution of pathogen virulence in IHNV, BCWD, and Perkinsus marinus.
Vaccination, Selective Breeding, and Virulence Evolution
Host vaccination and selective breeding for disease resistance are hypothesized to select for the evolution of increased pathogen virulence. This hypothesis comes from an extension of the virulence trade-off hypothesis. Take the vaccination case. Vaccines reduce host morbidity and mortality. Some vaccines however do not completely block pathogen infection (often termed leaky). Imagine a pathogen strain which is so virulent that in an unvaccinated host it kills the host before it can transmit, and thus has zero fitness. However, when the same virulent strain infects a vaccinated host it does not kill that host because the vaccine provides disease protection. If the vaccine was leaky, this virulent strain would be able to persist in the vaccinated host and potentially transmit to new hosts. Ultimately, vaccination allows for pathogen strains to maintain higher replication rates, and thus transmission rates, for less cost (host mortality reduction of transmission duration). This shifts the optimal level of virulence that provides the pathogen with maximum fitness upwards (Fig. 1). A similar logic could be made of selective breeding for disease resistance. Clearly, the evolution of virulence cannot occur if the vaccine or selective breeding regimes block pathogen transmission. The theory therefore hypothesizes only leaky regimes lead to increased virulence evolution. Furthermore, the costs of increased virulence evolution are only felt in unvaccinated or selectively breed hosts. This is likely to occur when the regimes are stopped, there is partial coverage, or transmission between treated and untreated populations occurs, such as farmed and wild hosts. We are exploring the impacts of vaccination and selective breeding on virulence evolution in IHNV and BCWD of salmonid fishes.
Multi-Species Pathogen Interactions
In nature, co-infections of hosts with multiple pathogen species are common. There are many ways that these pathogen species can interact, either directly through competition for resources such as host cells, or indirectly through processes such as the host immune response. This fascinating area of pathogen biology is rarely investigated. We explore the ecological interactions between multiple pathogen species that reside in the same host. We are particularly interested in how control practices against one species, might influence the ecology of other pathogen species in the system. We are currently conducting this research with IHNV and BCWD. Anecdotal evidence suggests there is an interaction between these two pathogens. It is common lore among fish farmers that the presence of one exacerbates the disease of the other. However, there is no published literature on this topic. We are investigating how co-infection, vaccination, and selective breeding will impact the replication, transmission, and fitness of these two pathogens. We are also eager to explore these topics in other systems.
Infectious hematopoietic necrosis virus (IHNV) is a single-stranded, negative-sense, RNA virus in the family Rhabdoviridae. The natural host of the virus is salmonid fishes, in which it typically causes an acute disease with mortality occurring between 7 and 20 days post exposure. Death is caused by necrosis of kidney and spleen. The virus is endemic in the Pacific Northwest of North America, where it is believed to have originated in sockeye salmon and then adapted via a host jump into cultured rainbow trout during the 1970s. It has since spread and become endemic and epidemic in trout farming in North America, and more recently Europe and Asia. IHNV is now considered one of the most significant sources of economic loss for the trout farm industry. A main research question for us is whether IHNV has increased in virulence since it emerged in trout farming. It is feasible, although inconclusive, that farming has impacted the evolution of the virus, given that the largest degree of IHNV genetic diversity is found in a 100km valley of the Snake River in Idaho where trout farming is highly concentrated. We have demonstrated substantial virulence variation between different IHHV isolates from the field. We explored how these virulence differences are associated with fitness differences. To do so we developed a real time quantitative PCR assay (qPCR) to distinguish and quantify genetically distinct field isolates (genotypes) which differ in virulence. Using this assay we quantify IHNV genotype fitness in experimental infections in live fish (in vivo). Through highly replicated in vivo experiments we discovered that IHNV virulence is positively associated with traits host entry, replication, competitive fitness, and shedding (Fig. 2). We are now exploring how these traits are associated with IHNV transmission success.
In addition to causing major trout farming production losses, IHNV also infects a variety of other salmonid species and threatens fisheries, hatcheries, and wild stock conservation and management. The transmission dynamics within and between wild and cultured salmonids are not well understood. We are currently exploring the epidemiology and transmission pathways of the virus in the Pacific Northwest of North America. We are interested in what role fitness and virulence might play in the emergence of new IHNV genotypes that have displaced previously dominant genotypes in a particular geographic region.
We are also exploring how vaccination and selective breeding for IHN disease resistance might impact the evolution of virulence. Selective breeding for IHN disease resistance has occurred for many decades in the trout farm industry with limited success. There has also been a global effort to develop vaccines to IHNV. DNA vaccines that are highly effective at protecting against IHN disease have been thoroughly described in published literature and a IHNV DNA vaccine is commercially available in Canada, where it has been used in Atlantic salmon culture since 2005. Little is known about what impact these practices are having on the evolution of the virus.
Bacterial cold water disease (BCWD) is caused by the bacteria Flavobacterium psychrophilum, which occurs worldwide in nearly all locations where salmonids are cultured. BCWD can present as an acute outbreak in small fish or as a chronic disease in larger fish, typically causing between 2-30% mortality. The disease causes significant economic losses in aquaculture production due to mortality and deformities in surviving fish as a result of tissue necrosis. Analysis of field isolates of BCWD has revealed substantial genetic diversity, and often multiple isolates are obtained from a single fish indicating mixed infection. It has been proposed that high virulence strains may be responsible for epizootics in fry. Current BCWD management frequently involves oral antibiotic therapy, chemical treatment, and/or a reduction in fish density. At present, there is no approved vaccine available for BCWD, although a live-attenuated vaccine is currently undergoing commercial evaluation. Other control strategies under evaluation include probiotic feeding and phage therapy. There have also been selective breeding efforts for BCWD resistance in rainbow trout by various groups in the field. Recently a selective breeding program was initiated at the USDA-ARS NCCCWA laboratory, and successfully developed a BCWD disease resistant trout line, which is undergoing field evaluation. We are investigating the association between BCWD virulence and fitness and what impact fish farming, vaccination, selective breeding, and co-infection might have on BCWD ecology and virulence evolution.
Hematodinium perezi is a syndinid dinoflagellate that infects the American blue crab, Callinectes sapidus, as well as several other crustaceans. The parasite lives in the hemolymph (blood) of the crab, where it rapidly proliferates, eventually killing the host. It was originally described in green crabs, Carcinus maenas, and velvet crabs Liocarcinus depurator from France. The parasite was first documented in blue crabs by Newman & Johnson and has a broad distribution on the eastern seaboard and Gulf coast of the USA. It is endemic in high salinity waters within this range and periodic outbreaks (prevalence >50%) cause significant crab mortalities. The parasite can be highly virulent in its blue crab host, showing mortality levels of up to 87% in naturally and experimentally infected blue crabs. Genetic diversity of H. perezi is known to be high in the field. We are exploring whether or not there is a genetic and phenotypic association between virulence and fitness for H. perezi. We are curious how this might shape the epidemiology of the parasite in the field.
The protistan parasite Perkinsus marinus infects the oyster Crassostrea virginica in estuaries from Maine to the Yucatán Peninsula. A drastic and lasting shift of increased virulence was observed in the P. marinus phenotype in the 1980s. This led to great intensification of P. marinus disease and dramatic diminution of oyster populations in Chesapeake Bay around 1986. Mean prevalence of P. marinus among oysters can now be greater than 90% where salinities are high, and the average parasitemia typically exceeds 10,000 parasite cells/host. It is hypothesized that a mechanism driving the phenotypic shift of P. marinus was adaptation to a host landscape greatly transformed by the emergence of Haplosporidium nelsoni (colloquially, “MSX”) a quarter century earlier. H. nelsoni is an invasive parasite introduced from Asia sometime prior to its emergence in Delaware Bay in 1957, and Chesapeake Bay in 1959. In 1959-1961 H. nelsoni killed > 90% of oysters in mesohaline waters inhabited by P. marinus. The virulence increase (i.e., the phenotype change) in P. marinus may therefore have been linked to H. nelsoni via the parasite’s genetic or environmental perturbation of the host populations. We are pursuing studies to determining the factors and mechanisms that caused the shift in P. marinus virulence. We are very curious how the seasonality peaks of this parasite during summer months might have played a role in its evolution.
Malaria, one of the world’s most deadly and infamous diseases, has a dynamic in-host ecology which heavily shapes its evolution. However, determining the epidemiological relevance of malaria in-host ecology has been hindered by an incomplete understanding of how it relates to the essential life history ingredient: between-host transmission. This lack of knowledge has potentially contributed to the ongoing failure to eradicate the disease. Our work has been to unravel the mysterious relationship between malaria in-host ecology and transmission to provide insights into effective disease control strategies.
Our research is conducted using the rodent malaria parasite Plasmodium chabaudi, which proves to be a highly malleable experimental and laboratory model for its human counterpart, Plasmodium falciparum. We developed a novel reverse transcription qPCR technique for clone specific assessment of P. chabaudi transmission stage (gametocyte) production. This was combined with current qPCR methods for quantification of in-host total parasite numbers, to examine in-host replication and transmission investment strategies in relation to infection stage and conspecific interactions.
This work revealed that the majority of malaria total parasite and transmission stage production occurs within the initial phase of infection. However, many individuals develop very low level long lasting chronic infections which are transmissible to vectors and likely to heavily impact the epidemiology of the parasite, confirming findings in the field. During the high parasite density stage at the start of the infection, in-host dynamics and conspecific competition were found to play a critical role in shaping the fitness of malaria genotypes co-infecting the same host. In response to conspecific competition, a small degree of transmission investment plasticity was observed, but this did not affect the competitive outcome.
Drug Resistance Evolution
Malaria has evolved resistance to virtually every drug that has been deployed in the field. As such, mixed infections of drug-resistant and sensitive malaria genotypes are known to be common. We explored the competitive dynamics between resistant and sensitive genotypes of P. chabaudi when they infect the same host, so as to better understand the ecology of drug resistance evolution. In the absence of drugs, resistance provides genotypes with a fitness cost of lower replication rates compared to sensitive genotypes. This allows for drug resistant genotypes to be competitively suppressed by sensitive genotypes in the absence of drugs. We found that strong drug dosage, which completely cleared the drug sensitive clone, released the resistant clone from competition, resulting in its enhanced growth and increased transmission potential. However, reducing drug dosages to levels that did not completely clear the sensitive genotype reduced the competitive release and enhanced growth of the resistant genotypes. Thus, drug dosages, which just alleviate clinical symptoms but do not completely clear parasites, will preserve within host competition and may reduce the rate at which drug resistance spreads (Fig. 3). This contradicts current medical thinking that incomplete drug treatment regimes will accelerate drug resistance evolution.