The Evolution of Virulence: A Unifying Link between Parasitology and Ecology

For most of the past century, the in parasitology and epidemiology has been that a parasite is a benign parasite (parasites being defined broadly to include multicellular, unicellular, and subcellular organisms). This view became firmly entrenched during the second quarter of this century (Smith, 1934; Swellengrebel, 1940). Although objections were raised during the middle decades of this century (Ball, 1943; Cockburn, 1963), they had little impact in the face of the more aesthetically appealing paradigm, which advocated evolution toward balanced, stable associations in which parasites succeeded through peaceful coexistence with their hosts. The problem with this view is that natural selection does not necessarily favor peaceful coexistence. Rather, it favors those genetic instructions that are passed on preferentially as a result of the characteristics they encode. If instructions that cause their bearer to exploit host resources more fully are more successfully passed into future generations as a result of that exploitation, then adaptation to the host will be associated with increased exploitation of the host. Insofar as increased exploitation causes increased harm, evolutionary adaptation to the host will result in increased harm to the host. This conclusion was reached independently by several investigators during the last quarter of the 20th century (Ewald, 1980, 1983; Price, 1980; Levin and Pimentel, 1981; Anderson and May, 1982)--in contrast to the situation during mid-century, this time an alternative theoretical framework grew out of this competing view. Now, at the end of the 20th century this competing paradigm has replaced the conventional wisdom among evolutionary biologists. This fundamental change in our view of host/parasite evolution promises to transform perspectives and research programs in parasitology and the health sciences. This transformation, however, is still in progress, largely because of the viscosity at the interface between disciplines. (For a more detailed history of these ideas, see Ewald [1994].) Even now the tenet that parasites are benign is considered by many to be a guiding principle. Parasitologists still often speak of benignness as a sign of a parasite or infer that mildness is a sign that a host is a natural host. Similarly, this outdated view has also colored the prevailing interpretations of aspects of infectious diseases, such as the origins and evolution of human immunodeficiency virus (HIV) (Ewald, 1994; Mindell et al., 1995). The inadequacy of this viewpoint can be seen by considering the evolutionary stability of a parasite that is peacefully coexisting with its host. Imagine such a well adapted parasite and ask whether a more exploitative variant of that parasite would be more or less successful. If that variant funneled the resources gained from this exploitation into its own propagation, would it pass on more copies of its genetic instructions than the less exploitative variant? For a multicellular parasite this increased exploitation might involve greater conversion of host tissues into parasite eggs or into parasite tissues, which would eventually foster greater reproduction. For a unicellular parasite or a virus, this extra exploitation might involve greater invasion of host tissues, liberating resources for the parasite's replication (as with Shigella), or it might involve production of a toxin that indirectly leads to an increased rate of propagation from a host, e.g., as in Vibrio cholerae or Corynebacterium diphtheriae (Ewald, 1994). If the extra exploitation involves no adverse effects on the host, then the variants that exploited the hosts more fully would win. By this line of reasoning, we expect natural selection to push exploitation at least to the point at which the parasite begins to impose negative effects on the host. The problem now becomes particularly interesting. If one agrees that natural selection could favor parasites that exploited hosts to this point, why not presume that it could favor parasites that exploited hosts even more intensely? Parasite variants that did so would still get the fitness benefits of that exploitation, but they may also incur fitness costs as a result of the negative effects of their exploitation on the host. Indeed, it is this tradeoff that is at the heart of current theory about the evolution of virulence (which I use broadly in this paper to mean the harmfulness of a consumer to the individual it is consuming). Current theory proposes that natural selection may favor even high levels of exploitation by parasites if the negative effects on the host have relatively small negative effects on parasite transmission. Countering these effects, of course, are the evolutionary responses of the host to the parasite. Hosts that reduce a parasite's exploitation leave more of their instructions. The end result is quite different from peaceful coexistence-it is an arms race between parasite and host, which is most intense among those parasites for which natural selection favors intense exploitation. This conclusion leads to an important question: What are the factors that drive some parasites toward intense exploitation of hosts and others toward mild coexistence? In this paper, I address this question and suggest that the answer will provide not only a clearer view of host/parasite coevolution, but of the entire spectrum of consumer associations. The approach promises not only to conceptually unify categories such as predation, parasitism, commensalism, and mutualism, but also to provide insight into the reasons for the evolution of complex life histories among parasites. Studies investigating the evolution of virulence have focused on host mobility because a reduction in mobility is one of the first discernable negative effects at the boundary between benignness and virulence, and because different parasites may depend on host mobility to different degrees. A reduction in host mobility should be especially costly for parasites that rely on host mobility for transmission. Conversely, if transmission is * This paper is based on the R. Barclay McGhee Lecture presented at the 60th Annual Meeting at Fort Collins, Colorado, in August 1994.