Evolution of parasite virulence: virulence-transmission trade-offs

Why do parasites - which rely on their hosts for survival - cause disease or even their hosts? A major evolutionary theory poses that parasite virulence can evolve as an unavoidable consequence of natural selection on parasite transmission. The idea is that parasites need to transmit between hosts and that higher host exploitation results in greater transmission. However, parasite growth also uses up host resources and damages tissues, and thereby results in disease and an increased probability that the host will die. Thus, high parasite growth inside a host has costs and benefits: the costs are increased disease and mortality, while the benefits are increased transmission to new hosts. On the basis of this trade-off it is expected that parasite strains with intermediate growth rates are selected for in nature: this is because these parasites grow fast enough to obtain transmission to new hosts, but not so fast as to kill the host before such transmission can occur.

We have experimentally tested this theory, using the protozoan parasite Ophryocystis elektroscirrha, which infects monarch butterflies (Danaus plexippus) in natural populations. This parasite infects monarch butterflies at the larval stage, when caterpillars ingest parasite spores that are deposited by adult butterflies on the milkweed plants that these caterpillars eat. Parasites replicate to high numbers in the monarch caterpillar, such that adult butterflies can carry millions of parasite spores. In our work, we found that these high numbers of parasite spores are necessary to ensure parasite transmission to the next generation: transmission of spores occurs from egg-laying adult butterflies to their offspring caterpillars.

However, parasite growth is also detrimental to the monarch, and monarchs that carry high numbers of spores are more likely to die before the adult stage, and mate less well and live shorter as adults. Interestingly, these results imply that very high spore production is also bad for the parasite, which requires the monarch host to survive, mate and lay eggs. Our work showed that parasite fitness was highest at intermediate numbers of parasite spores, where the costs of high numbers (virulence) were balanced by the benefits (transmission to the next generation of hosts). This work suggests that parasite virulence may indeed evolve as a consequence of selection on parasite transmission: by selecting for parasites that can transmit between hosts, nature simultaneously selects for parasites that cause disease to their hosts.  
Left: parasite spores on monarch butterfly abdominal scales. Right: parasite fitness is highest at an intermediate growth inside the host due to a trade-off between virulence and transmission.



Evolution of parasite virulence: within-host competition

A second way in which parasite virulence may be selected for is through within-host competition. Although unknown to most people, many to most parasite infections consist not of one, but multiple strains of the same parasite species. As a result, these parasite strains face competition for the limited resources that the single host can provide. Given such competition, it has often been predicted that parasites that grow faster will be better competitors and have a higher probability to transmit to new hosts. Because parasites with higher growth rates also cause more virulence, such competition would select for parasites that cause greater disease and are more likely to kill the host.

We tested this hypothesis using laboratory mice and the rodent malaria parasite Plasmodium chabaudi. We infected mice with either single parasite strains, or mixes of parasite strains, and then tracked parasite growth over time. We also tested whether parasites that were better at competing inside the mouse host also had a higher probability to infect malaria mosquitoes. Our results showed that parasites that had higher within-host growth caused more disease to their mouse hosts (i.e. they caused greater anemia and more weight loss). These strains were also better at competing in mixed-strain infections, and also were more likely to infect mosquitoes from such mixed infections. These results suggest that the competition that occurs in mixed-strain infections favors more virulent parasites, and thereby selects for the evolution of higher virulence in a population.

Malaria parasites that cause more virulence have higher competitive ability and obtain greater transmission to mosquitoes.

Effects of environmental factors on parasite virulence evolution



Parasites and their hosts do not live in isolation. Instead, they are part of much larger communities of interacting species and environmental variables. Part of our work focuses on understanding how such species and environmental factors affect parasite virulence and transmission. In particular, monarch butterflies use species of milkweed as their larval food plants, and we have found that some milkweed species can reduce parasite infection and growth and thereby relieve the disease symptoms of infected monarch butterflies. Thus, some milkweeds are medicinal to monarch butterflies. In collaboration with Mark Hunter’s laboratory at the University of Michigan, we have also shown that these medicinal effects are likely due to the toxic chemicals that these milkweeds contain. These results are significant, because they demonstrate that the environment in which hosts and parasites interact can crucially affect parasite virulence.

Our theoretical work has so far shown that medicinal milkweeds can select for more virulent parasites, since these milkweeds reduce parasite growth to sub-optimal levels, resulting in selection for parasites with higher intrinsic growth rate. We are currently testing the hypothesis that medicinal milkweeds in nature also select for higher virulence in monarch parasites. To this end, we collect parasites from multiple populations to compare their intrinsic levels of virulence. We are also using molecular markers to test whether the populations of monarchs that we study are indeed distinct populations.

This monarch caterpillar feeds on Asclepias tuberosa (butterfly weed). In contrast to some other species of milkweed, A. tuberosa contains low concentrations of toxic chemicals, and does not provide protection to the monarch butterfly.

Ecology and evolution of self-medication

Because parasites cause disease and death in their hosts, there should be strong selection for hosts to evolve ways to protect themselves against parasites. One of the most intriguing ways in which animals can protect themselves against disease is medication, through which animals use naturally available drugs to fight their parasites. Such medication has often been suggested to occur in great apes and other primates, but there remains a lack of strong evidence for medication in nature. Our lab recently tested for the existence of medication in monarch butterflies. Based on our findings that certain milkweed species can reduce parasite infection and growth and thereby relieve disease symptoms, we predicted that monarch butterflies would be able to specifically use such medicinal milkweed plants when infected. In a series of experiments, we found that monarch caterpillars cannot actively choose medicinal plants when infected. However, we found that infected female butterflies preferentially lay their eggs on milkweed that will make their offspring less sick. These results suggest that monarch butterflies have evolved the ability to medicate their offspring, and provide strong evidence that wild animals can use medication to fight their parasitic infections. Together with Mark Hunter’s lab at the University of Michigan we are working to identify the chemicals that make milkweeds medicinal. Because many parasites afflicting humans – including those that cause malaria, toxoplasmosis and cryptosporidiosis – are closely related to the parasites afflicting monarch butterflies, identifying these chemicals may provide new drugs against human parasites.

As part of this work, we are currently testing the hypothesis that levels of disease risk determine the type of medication that animals evolve. In particular, we predict that when disease risk is high and predictable, animals will evolve a prophylactic form of medication ("take drugs before getting sick"), and that when disease risk is low, animals will evolve a therapeutic form of self-medication ("only take drugs when sick").


When given the choice between the medicinal milkweed Asclepias curassavica and the non-medicinal A. incarnata, monarch caterpillars have no preference for either species (left). However, infected adult female monarchs preferentially lay their eggs on the medicinal A. curassavica (right), which results in lower infection rates and less disease in their offspring.



Evolution of host defense against parasite infection

Self-medication is just one of the many ways in which organisms may protect themselves against parasites. Other ways involve resistance (prevent infection or reduce parasite numbers) and tolerance (alleviate the symptoms of infection without preventing infection or reducing parasite numbers). We have studied both these mechanisms in monarch butterflies and found that monarchs vary genetically in resistance, but not tolerance. To better understand monarch defenses against parasite infection, we are currently working on a project to sequence the transcriptome of infected and uninfected butterflies. This will enable us to identify genes that are specifically expressed in monarchs infected with their parasites. We are doing this work in collaboration with Nicole Gerardo and Tim Read's Emory GRA Genome Center, both here at Emory.



Ecology and evolution of infections of multiple parasite species


Parasite infections in nature often consist of multiple parasite species. Such co-infections are important since they can have strong effects on the virulence and transmission of all parasite species involved.

We are studying co-infections of the protozoan parasite Ophryocystis elektroscirrha and the fly parasitoid Lespesia archippivora, two parasites that co-occur in natural populations of monarch butterflies. So far, we have found that the virulent protozoan can reduce the mortality caused by the lethal parasitoid fly, and can thereby provide some protection to the monarch host. We are currently investigating the mechanisms by which this protection may arise, focusing on two alternative immunological mechanisms: first, the protozoan parasite may increase the monarch's immune response against the parasitoid fly; and second, the protozoan parasite may reduce the immunopathology induced by fly infection.



A monarch caterpillar with 3 parasitoid fly maggots.

Evolution of virulence in Varroa mites of honeybees

We recently started to work with Berry Brosi (Emory) and Keith Delaplane (University of Georgia) to study whether current bee-keeping practices select for highly virulent bee mites. Honeybees are declining worldwide, and much of these declines are attributed to "Colony Collapse Disorder". However, Varroa mites pose a greater threat to honeybees, causing great natural and economic losses worldwide by wiping out large fractions of beekeeper' colonies. One explanation for the great harm done by these mites is that current bee-keeping practices are actually selecting for highly virulent mites by facilitating high levels of mite transmission between colonies, apiaries and different geographic regions. In theory, such facilitation of transmission should select for highly virulent parasites. We are currently testing this hypothesis by comparing the transmission and virulence of mites in managed and feral bee colonies.


Parasitoid avoidance in fruit flies

Like other organisms, fruit flies have many natural enemies, including parasitoid wasps that lay their eggs inside fruit fly larvae. Our colleague Todd Schlenke has studied these wasps for many years, and we have recently joined his research by studying whether fruit flies can behaviorally avoid wasp attacks. In particular, we are investigating whether fruit flies lay fewer eggs when surrounded by parasitoid wasps, and whether they can postpone their egg-laying until wasp threats have subsided. We are also investigating the mechanisms by which fruit flies detect the presence of wasps, using all sorts of cool mutants that will probably never be available for monarch butterflies.



Effects of competition between drug-resistant and –sensitive malaria parasites on the spread of drug resistance

As outlined above, parasite infections often consist of multiple parasite genotypes. In many cases, such mixed infections can contain parasites that are sensitive and those that are resistant to anti-parasitic drugs. Because parasites in mixed infections have to compete with each other for limited resources (see above), it can be expected that drug-resistant strains do not produce as many offspring and do not obtain as much transmission when competing with other strains than if they infected a host alone. Interestingly, if this is the case, then it can also be expected that the removal of drug-sensitive strains through the use of drugs removes this competition. This, in turn, could result in an increase in growth and transmission of the drug-resistant strain, and thereby aid in its spread through a population.

Using the rodent malaria parasite Plasmodium chabaudi in laboratory mice, we carried out a number of experiments that confirm this notion: drug-resistant strains suffer from competition by drug-sensitive strains; and when these sensitive strains are removed with drugs, the resistant strain grows to high numbers and infects more mosquitoes. Indeed, with our colleagues at the University of Edinburgh we found that in some cases the resistant strains may grow to even higher numbers than they would have done on their own. In contrast, in experiments where we used sub-curative drug doses (which reduced but did not remove the drug-sensitive strain), we found that drug-resistant parasites still suffered from competition, and did not grow to high numbers inside the host. These results are significant, because they suggest that the removal of competition through drugs greatly contributes to the spread of drug resistance. This work also suggests that sub-curative drug treatment – which reduces disease symptoms, but does not clear infections – could aid in slowing down the spread of drug resistance. We are currently exploring ways in which to follow up this work in human malaria.


Mouse red blood cells infected with malaria parasites.