Shared control of epidemiological traits in a coevolutionary model of host-parasite interactions

Most models concerning the evolution of a parasite's virulence and its host's resistance assume that each component of the relationship (transmission, virulence, recovery, etc.) is controlled by either the host or the parasite but not by both. We present a model that describes the coevolution of host and parasite, assuming that the rate of transmission or the virulence depends on both genotypes. The evolution of these traits is constrained by trade-offs that account for costs of defense and attack strategies, in line with previous studies on the separate evolution of the host and the parasite. Considering shared control by the host and the parasite in determining the traits of the relationship leads to several novel predictions. First, the host should evolve maximal investment in defense against parasites with an intermediate replication rate. Second, the evolution of the parasite strongly depends on the way the host's defense is described. Third, the coevolutionary process may lead to decreasing the parasite's virulence as a response to a rise in the host's background mortality, contrary to classical predictions.     

Spatial structure,host heterogeneity and parasite virulence: implications for vaccine‐driven evolution

Natural host‐parasite interactions exhibit considerable variation in host quality, with profound consequences for disease ecology and evolution. For instance, treatments (such as vaccination) may select for more transmissible or virulent strains. Previous theory has addressed the ecological and evolutionary impact of host heterogeneity under the assumption that hosts and parasites disperse globally. Here, we investigate the joint effects of host heterogeneity and local dispersal on the evolution of parasite life‐history traits. We first formalise a general theoretical framework combining variation in host quality and spatial structure. We then apply this model to the specific problem of parasite evolution following vaccination. We show that, depending on the type of vaccine, spatial structure may select for higher or lower virulence compared to the predictions of non‐spatial theory. We discuss the implications of our results for disease management, and their broader fundamental relevance for other causes of host heterogeneity in nature.     

Evolution of spatially structured host–parasite interactions

Spatial structure has dramatic effects on the demography and the evolution of species. A large variety of theoretical models have attempted to understand how local dispersal may shape the coevolution of interacting species such as host–parasite interactions. The lack of a unifying framework is a serious impediment for anyone willing to understand current theory. Here, we review previous theoretical studies in the light of a single epidemiological model that allows us to explore the effects of both host and parasite migration rates on the evolution and coevolution of various life‐history traits. We discuss the impact of local dispersal on parasite virulence, various host defence strategies and local adaptation. Our analysis shows that evolutionary and coevolutionary outcomes crucially depend on the details of the host–parasite life cycle and on which life‐history trait is involved in the interaction. We also discuss experimental studies that support the effects of spatial structure on the evolution of host–parasite interactions. This review highlights major similarities between some theoretical results, but it also reveals an important gap between evolutionary and coevolutionary models. We discuss possible ways to bridge this gap within a more unified framework that would reconcile spatial epidemiology, evolution and coevolution.     

Coevolution between parasite virulence and host life-history traits

Epidemiological models generally explore the evolution of parasite life-history traits, namely, virulence and transmission, against a background of constant host life-history traits. However, life-history models have predicted the evolution of host traits in response to parasitism. The coevolution of host and parasite life-history traits remains largely unexplored. We present an epidemiological model, based on resource allocation theory, that provides an analysis of the coevolution between host reproductive effort and parasite virulence. This model allows for hosts with either a fixed (i.e., genetic) or conditional (i.e., a phenotypically plastic) response to parasitism. It also considers superinfections. We show that parasitism always favors increased allocation to host reproduction, but because of epidemiological feedbacks, the evolutionarily stable host reproductive effort does not always increase with parasite virulence. Superinfection drives the evolution of parasite virulence and acts on the evolution of the host through parasite evolution, generally leading to higher host reproductive effort. Coevolution, as opposed to cases where only one of the antagonists evolves, may generate correlations between host and parasite life-history traits across environmental gradients affecting the fecundity or the survival of the host. Our results provide a theoretical framework against which experimental coevolution outcomes or field observations can be contrasted.     

Virulence and competitive ability in an obligately killing parasite

Mixed infections are thought to have a major influence on the evolution of parasite virulence. During a mixed infection, higher within‐host parasite growth is favored under the assumption that it is critical to the competitive success of the parasite. As within‐host parasite growth may also increase damage to the host, a positive correlation is predicted between virulence and competitive success. However, when parasites must kill their hosts in order be transmitted, parasites may spend energy on directly attacking their host, even at the cost of their within‐host growth. In such systems, a negative correlation between virulence and competitive success may arise. We examined virulence and competitive ability in three sympatric species of obligately killing nematode parasites in the genus Steinernema. These nematodes exist in a mutualistic symbiosis with bacteria in the genus Xenorhabdus. Together the nematodes and their bacteria kill the insect host soon after infection, with reproduction of both species occurring mainly after host death. We found significant differences among the three nematode species in the speed of host killing. The nematode species with the lowest and highest levels of virulence were associated with the same species of Xenorhabdus, indicating that nematode traits, rather than the bacterial symbionts, may be responsible for the differences in virulence. In mixed infections, host mortality rate closely matched that associated with the more virulent species, and the more virulent species was found to be exclusively transmitted from the majority of coinfected hosts. Thus, despite the requirement of rapid host death, virulence appears to be positively correlated with competitive success in this system. These findings support a mechanistic link between parasite growth and both anti‐competitor and anti‐host factors.     

Within‐host interactions shape virulence‐related traits of trematode genotypes

Within‐host interactions between co‐infecting parasites can significantly influence the evolution of key parasite traits, such as virulence (pathogenicity of infection). The type of interaction is expected to predict the direction of selection, with antagonistic interactions favouring more virulent genotypes and synergistic interactions less virulent genotypes. Recently, it has been suggested that virulence can further be affected by the genetic identity of co‐infecting partners (G × G interactions), complicating predictions on disease dynamics. Here, we used a natural host–parasite system including a fish host and a trematode parasite to study the effects of G × G interactions on infection virulence. We exposed rainbow trout (Oncorhynchus mykiss) either to single genotypes or to mixtures of two genotypes of the eye fluke Diplostomum pseudospathaceum and estimated parasite infectivity (linearly related to pathogenicity of infection, measured as coverage of eye cataracts) and relative cataract coverage (controlled for infectivity). We found that both traits were associated with complex G × G interactions, including both increases and decreases from single infection to co‐infection, depending on the genotype combination. In particular, combinations where both genotypes had low average infectivity and relative cataract coverage in single infections benefited from co‐infection, while the pattern was opposite for genotypes with higher performance. Together, our results show that infection outcomes vary considerably between single and co‐infections and with the genetic identity of the co‐infecting parasites. This can result in variation in parasite fitness and consequently impact evolutionary dynamics of host–parasite interactions.     

Parasite variation and the evolution of virulence in a Daphnia-microparasite system

Understanding genetic relationships amongst the life-history traits of parasites is crucial for testing hypotheses on the evolution of virulence. This study therefore examined variation between parasite isolates (the bacterium Pasteuria ramosa) from the crustacean Daphnia magna. From a single wild-caught infected host we obtained 2 P. ramosa isolates that differed substantially in the mortality they caused. Surprisingly, the isolate causing higher early mortality was, on average, less successful at establishing infections and had a slower growth rate within hosts. The observation that within-host replication rate was negatively correlated with mortality could violate a central assumption of the trade-off hypothesis for the evolution of virulence, but we discuss a number of caveats which caution against premature rejection of the trade-off hypothesis. We sought to test if the characteristics of these parasite isolates were constant across host genotypes in a second experiment that included 2 Daphnia host clones. The relative growth rates of the two parasite isolates did indeed depend on the host genotype (although the rank order did not change). We suggest that testing evolutionary hypotheses for virulence may require substantial sampling of both host and parasite genetic variation, and discuss how selection for virulence may change with the epidemiological state of natural populations and how this can promote genetic variation for virulence.     

Host mortality, predation and the evolution of parasite virulence

One of the most accepted views in the theoretical literature on virulence evolution is that a parasite's virulence will evolve to higher levels when its host's background mortality rate increases. Surprisingly, however, although many sources of background mortality involve predation, there has not yet been any theoretical research that explicitly considers how the dynamics of this important ecological interaction affects virulence evolution. Here, we consider how predation affects virulence evolution by explicitly introducing a predator into a classical susceptible–infected–susceptible epidemiological model. We find that, contrary to previous predictions, different sources of host mortality affect virulence evolution in different ways. Moreover, the way in which virulence evolution is affected depends on how tightly coupled the predator's dynamics are to the host population, and this can result in somewhat counterintuitive results. For example, indirect ecological effects can cause elevated host mortality to result in the evolution of lower parasite virulence, even if this elevated mortality arises from factors unrelated to predation.     

Life history and virulence are linked in the ectoparasitic salmon louse Lepeophtheirus salmonis

Models of virulence evolution for horizontally transmitted parasites often assume that transmission rate (the probability that an infected host infects a susceptible host) and virulence (the increase in host mortality due to infection) are positively correlated, because higher rates of production of propagules may cause more damages to the host. However, empirical support for this assumption is scant and limited to microparasites. To fill this gap, we explored the relationships between parasite life history and virulence in the salmon louse, Lepeophtheirus salmonis, a horizontally transmitted copepod ectoparasite on Atlantic salmon Salmo salar. In the laboratory, we infected juvenile salmon hosts with equal doses of infective L. salmonis larvae and monitored parasite age at first reproduction, parasite fecundity, area of damage caused on the skin of the host, and host weight and length gain. We found that earlier onset of parasite reproduction was associated with higher parasite fecundity. Moreover, higher parasite fecundity (a proxy for transmission rate, as infection probability increases with higher numbers of parasite larvae released to the water) was associated with lower host weight gain (correlated with lower survival in juvenile salmon), supporting the presence of a virulence–transmission trade‐off. Our results are relevant in the context of increasing intensive farming, where frequent anti‐parasite drug use and increased host density may have selected for faster production of parasite transmission stages, via earlier reproduction and increased early fecundity. Our study highlights that salmon lice, therefore, are a good model for studying how human activity may affect the evolution of parasite virulence.     

How parasite interaction strategies alter virulence evolution in multi‐parasite communities

The majority of organisms host multiple parasite species, each of which can interact with hosts and competitors through a diverse range of direct and indirect mechanisms. These within‐host interactions can directly alter the mortality rate of coinfected hosts and alter the evolution of virulence (parasite‐induced host mortality). Yet we still know little about how within‐host interactions affect the evolution of parasite virulence in multi‐parasite communities. Here, we modeled the virulence evolution of two coinfecting parasites in a host population in which parasites interacted through cross immunity, immune suppression, immunopathology, or spite. We show (1) that these within‐host interactions have different effects on virulence evolution when all parasites interact with each other in the same way versus when coinfecting parasites have unique interaction strategies, (2) that these interactions cause the evolution of lower virulence in some hosts, and higher virulence in other hosts, depending on the hosts infection status, and (3) that for cross immunity and spite, whether parasites increase or decrease the evolutionarily stable virulence in coinfected hosts depended on interaction strength. These results improve our understanding of virulence evolution in complex parasite communities, and show that virulence evolution must be understood at the community scale.     

A UNIFIED MODEL FOR THE COEVOLUTION OF RESISTANCE,TOLERANCE, AND VIRULENCE

We present a general host–parasite model that unifies previous theory by investigating the coevolution of virulence, resistance, and tolerance, with respect to multiple physiological, epidemiological, and environmental parameters. Four sets of new predictions emerge. First, compared to virulence coevolving with resistance or tolerance, three‐trait coevolution promotes more virulence and less tolerance, and broadens conditions under which pure defenses evolve. Second, the cost and efficiency of virulence and the epidemiological rates are the key factors of virulence coevolving with resistance and tolerance. Maximum virulence evolves for intermediate infection rate, at which coevolved levels of resistance and tolerance are both high. The influence of host and parasite background mortalities is strong on the evolution of defenses and weak on the coevolution of virulence. Third, evolutionary correlations between defenses can switch sign along single‐parameter gradients. The evolutionary trade‐off between resistance and tolerance may coevolve with virulence that either increases or decreases monotonically, depending on the underlying parameter gradient. Fourth, despite global attractiveness and stability of coevolutionary equilibria, not‐so‐rare and not‐so‐small mutations can beget large variation in virulence and defenses around equilibrium, in the form of transient “evolutionary spikes.” Implications for evolutionary management of infections are discussed and directions for future research are outlined.     

HOST–PARASITE GENETIC INTERACTIONS AND VIRULENCE-TRANSMISSION RELATIONSHIPS IN NATURAL POPULATIONS OF MONARCH BUTTERFLIES

Evolutionary models predict that parasite virulence (parasite-induced host mortality) can evolve as a consequence of natural selection operating on between-host parasite transmission. Two major assumptions are that virulence and transmission are genetically related and that the relative virulence and transmission of parasite genotypes remain similar across host genotypes. We conducted a cross-infection experiment using monarch butterflies and their protozoan parasites from two populations in eastern and western North America. We tested each of 10 host family lines against each of 18 parasite genotypes and measured virulence (host life span) and parasite transmission potential (spore load). Consistent with virulence evolution theory, we found a positive relationship between virulence and transmission across parasite genotypes. However, the absolute values of virulence and transmission differed among host family lines, as did the rank order of parasite clones along the virulence-transmission relationship. Population-level analyses showed that parasites from western North America caused higher infection levels and virulence, but there was no evidence of local adaptation of parasites on sympatric hosts. Collectively, our results suggest that host genotypes can affect the strength and direction of selection on virulence in natural populations, and that predicting virulence evolution may require building genotype-specific interactions into simpler trade-off models.     

THE INFLUENCE OF HOST DEMOGRAPHY ON THE EVOLUTION OF VIRULENCE OF A MICROSPORIDIAN GUT PARASITE

It is predicted that host exploitation should evolve to maximize parasite fitness and that virulence (= parasite-induced host mortality) evolves along with the rate of host exploitation. If the life expectancy of a parasite is short, it is expected to evolve a higher rate of host exploitation and therefore higher virulence because the penalty to the parasite for killing the host is reduced. We tested this hypothesis by keeping for 14 months the horizontally transmitted microsporidian parasite Glugoides intestinalis in mono-clonal host cultures (Daphnia magna) under conditions of high and low host background mortality. High host mortality, and thus parasite mortality, was achieved by replacing weekly 70–80% of all hosts in a culture with uninfected hosts from stock cultures (Replacement lines). In the low-mortality treatment no replacement took place. Contrary to our expectation, parasites from the Replacement lines evolved a lower within-host growth rate and virulence than parasites from the Nonreplacement lines. Across lines we found a strong positive correlation between within-host growth rate and virulence. We did further experiments to answer the question why our data did not support the predictions. Sporophorous vesicles (SVs, spore clusters) were smaller in doubly infected than in singly infected host-gut cells, indicating that competition within cells bears costs for the parasite. Due to our experimental protocol, the average life span of infections had been much higher in the Nonreplacement lines. Since the number of parasites inside a host increases with the time since infection, long-lasting infections led to high frequencies of multiply infected host-gut cells. Therefore, we speculated that within-cell competition was more severe in the Nonreplacement lines and may have led to selection for accelerated within-host growth. SVs in the Nonreplacement lines were indeed significantly larger. Our results point out that single-factor explanations for the evolution of virulence can lead to wrong predictions and that multiple infections are an important factor in virulence evolution.     

Evolution of tolerance: the genetic basis of a host's resistance against parasite manipulation

Despite considerable theoretical advances in the evolutionary biology of host–parasite systems, our knowledge of host–parasite coevolution in natural systems is often limited. Among the reasons for the lag of experimental insight behind theory is that the parasite's virulence is not a simple trait that is controlled by the parasite's genes. Rather, virulence can be expressed in several traits due to the subtle interactions between the host and the parasite. Furthermore, the host might evolve tolerance to the parasite if there is sufficient genetic variance to reduce the detrimental effect of the parasite on these traits. We studied the traits underlying virulence and the genetic potential to evolve tolerance to infection in the host–parasite system Aedes aegypti – Brachiola algerae . We reared the mosquitoes in a half-sib design, exposed half of the individuals in each full-sib family to the parasite and measured several life history traits – juvenile mortality, age at pupation and adult size – of infected and uninfected individuals. Virulence was due in large part to a delay of the mosquito's age at pupation by about 10%. Although this imposes strong selection pressure on the mosquito to resist the parasite, all of the mosquitoes were infected, implying a lack of resistance. Furthermore, although additive genetic variance was present for other traits, we found no indication of additive genetic variation for the age at pupation, nor for the delay of pupation due to infection, implying no potential for the evolution of tolerance. Overall, the results suggest that in this host–parasite system, the host has little evolutionary control over the expression of the parasite's virulence.     

The impact of transmission mode on the evolution of benefits provided by microbial symbionts

While past work has often examined the effects of transmission mode on virulence evolution in parasites, few studies have explored the impact of horizontal transmission on the evolution of benefits conferred by a symbiont to its host. Here, we identify three mechanisms that create a positive covariance between horizontal transmission and symbiont‐provided benefits: pleiotropy within the symbiont genome, partner choice by the host, and consumption of host waste by‐products by symbionts. We modify a susceptible‐infected model to incorporate the details of each mechanism and examine the evolution of symbiont benefits given variation in either the immigration rate of susceptible hosts or the rate of successful vertical transmission. We find conditions for each case under which greater opportunity for horizontal transmission (higher migration rate) favors the evolution of mutualism. Further, we find the surprising result that vertical transmission can inhibit the evolution of benefits provided by symbionts to hosts when horizontal transmission and symbiont‐provided benefits are positively correlated. These predictions may apply to a number of natural systems, and the results may explain why many mutualisms that rely on partner choice often lack a mechanism for vertical transmission.     

Intensive Farming: Evolutionary Implications for Parasites and Pathogens

An increasing number of scientists have recently raised concerns about the threat posed by human intervention on the evolution of parasites and disease agents. New parasites (including pathogens) keep emerging and parasites which previously were considered to be 'under control' are re-emerging, sometimes in highly virulent forms. This re-emergence may be parasite evolution, driven by human activity, including ecological changes related to modern agricultural practices. Intensive farming creates conditions for parasite growth and transmission drastically different from what parasites experience in wild host populations and may therefore alter selection on various traits, such as life-history traits and virulence. Although recent epidemic outbreaks highlight the risks associated with intensive farming practices, most work has focused on reducing the short-term economic losses imposed by parasites, such as application of chemotherapy. Most of the research on parasite evolution has been conducted using laboratory model systems, often unrelated to economically important systems. Here, we review the possible evolutionary consequences of intensive farming by relating current knowledge of the evolution of parasite life-history and virulence with specific conditions experienced by parasites on farms. We show that intensive farming practices are likely to select for fast-growing, early-transmitted, and hence probably more virulent parasites. As an illustration, we consider the case of the fish farming industry, a branch of intensive farming which has dramatically expanded recently and present evidence that supports the idea that intensive farming conditions increase parasite virulence. We suggest that more studies should focus on the impact of intensive farming on parasite evolution in order to build currently lacking, but necessary bridges between academia and decision-makers.     

THE EVOLUTIONARY IMPLICATIONS OF CONFLICT BETWEEN PARASITES WITH DIFFERENT TRANSMISSION MODES

Understanding the processes that shape the evolution of parasites is a key challenge for evolutionary biology. It is well understood that different parasites may often infect the same host and that this may have important implications to the evolutionary behavior. Here we examine the evolutionary implications of the conflict that arises when two parasite species, one vertically transmitted and the other horizontally transmitted, infect the same host. We show that the presence of a vertically transmitted parasite (VTP) often leads to the evolution of higher virulence in horizontally transmitted parasites (HTPs), particularly if the VTPs are feminizing. The high virulence in some HTPs may therefore result from coinfection with cryptic VTPs. The impact of an HTP on a VTP evolution depends crucially on the nature of the life‐history trade‐offs. Fast virulent HTPs select for intermediate feminization and virulence in VTPs. Coevolutionary models show similar insights, but emphasize the importance of host life span to the outcome, with higher virulence in both types of parasite in short‐lived hosts. Overall, our models emphasize the interplay of host and parasite characteristics in the evolutionary outcome and point the way for further empirical study.     

Pesticide exposure strongly enhances parasite virulence in an invertebrate host model

Parasites are a common and constant threat to organisms at all levels of phyla. The virulence of a parasite, defined as the impact on survival and reproduction of its host, depends on the specific host–parasite combination and can also be influenced by environmental conditions. Environmental pollution might be an additional factor influencing host–parasite interactions. We here aimed to test whether the combined stress of pollutant exposure and parasite challenge results in stronger impacts on host organisms than expected from the single stressors applied alone. We used the water flea Daphnia magna and two of its endoparasites, the bacterium Pasteuria ramosa and the microsporidium Flabelliforma magnivora, as invertebrate host–parasite models. For each parasite, we tested in a full‐factorial design for interactions between parasitism and pollution using the neurotoxic pesticide carbaryl as a model substance. Sublethal concentrations of the pesticide synergistically enhanced the virulence of both parasites by increasing host mortality. Furthermore, host castration induced by P. ramosa was accelerated by carbaryl exposure. These effects likely reflect decreased host resistance due to direct or indirect immunosuppressive activity of carbaryl. The present study provides experimental evidence that the in vivo development of infectious diseases can be influenced by a pesticide at environmentally realistic concentrations. This implies that host–parasite interactions and subsequently co‐evolution might be influenced by environmental pollution at toxicant concentrations being sublethal to parasite‐free hosts. Standard toxicity testing as employed in the current way of conducting ecological risk assessments for anthropogenic substances does not consider natural antagonists such as infectious diseases, and thereby likely underestimates the impact these substances may pose to natural populations in the environment.     

Host-parasite interactions for virulence and resistance in a malaria model system

A rich body of theory on the evolution of virulence (disease severity) attempts to predict the conditions that cause parasites to harm their hosts, and a central assumption to many of these models is that the relative virulence of pathogen strains is stable across a range of host types. In contrast, a largely nonoverlapping body of theory on coevolution assumes that the fitness effects of parasites on hosts is not stable across host genotype, but instead depends on host genotype by parasite genotype interactions. If such genetic interactions largely determine virulence, it becomes difficult to predict the strength and direction of selection on virulence. In this study, we tested for host-by-parasite interactions in a medically relevant vertebrate disease model: the rodent malaria parasite Plasmodium chabaudi in laboratory mice. We found that parasite and particularly host main effects explained most of the variance in virulence (anaemia and weight loss), resistance (parasite burden) and transmission potential. Host-by-parasite interactions were of limited influence, but nevertheless had significant effects. This raises the possibility that host heterogeneity may affect the rate of any parasite response to selection on virulence. This study of rodent malaria is one of the first tests for host-by-parasite interactions in any vertebrate disease; host-by-parasite interactions typical of those assumed in coevolutionary models were present, but were by no means pervasive.