What Can Phages Tell Us about Host-Pathogen Coevolution?

Abstract

The outcomes of host-parasite interactions depend on the coevolutionary forces acting upon them, but because every host-parasite relation is enmeshed in a web of biotic and abiotic interactions across a heterogeneous landscape, host-parasite coevolution has proven difficult to study. Simple laboratory phage-bacteria microcosms can ameliorate this difficulty by allowing controlled, well-replicated experiments with a limited number of interactors. Genetic, population, and life history data obtained from these studies permit a closer examination of the fundamental correlates of host-parasite coevolution. In this paper, I describe the results of phage-bacteria coevolutionary studies and their implications for the study of host-parasite coevolution. Recent experimental studies have confirmed phage-host coevolutionary dynamics in the laboratory and have shown that coevolution can increase parasite virulence, specialization, adaptation, and diversity. Genetically, coevolution frequently proceeds in a manner best described by the Gene for Gene model, typified by arms race dynamics, but certain contexts can result in Red Queen dynamics according to the Matching Alleles model. Although some features appear to apply only to phage-bacteria systems, other results are broadly generalizable and apply to all instances of antagonistic coevolution. With laboratory host-parasite coevolutionary studies, we can better understand the perplexing array of interactions that characterize organismal diversity in the wild.

Figure 1

Fitnesses are shown for four parasite genotypes on each host genotype as implied by MA (left) and GFG (right) coevolutionary models. In the MA model, a parasite's genotype must precisely match a host's genotype in order to avoid recognition by the host's immune system and reproduce in the host. One consequence is that the number of parasite alleles matches the number of host alleles. By contrast, in the GFG model, a host is susceptible to all parasites except those for which it has a corresponding resistance allele. In this scenario, parasite alleles can outnumber host alleles. Figure modified from [42, 44].

Figure 2

Plots represent allele frequency changes over time as predicted by GFG (a) and MA (b) models of coevolution. In the GFG model, directional selection fixes host (blue) and parasite (red) alleles arising via mutation. Each specific host resistance allele interacts with a specific parasite avirulence gene. Parasites counter host resistance via mutations in avirulence genes. Over time, genetic changes accumulate in both populations. By contrast, virulence and resistance alleles persist as dynamic polymorphisms in the MA models. Here parasites (red) become specialized for a common host genotype (blue), reducing its fitness. Over time, this host genotype will decline in frequency as less common genotypes are favored because of reduced parasite load. Reduced host frequency reduces the benefits of parasite specialization on this host relative to more common host genotypes. Reduced parasite load leads to increased host fitness, causing the cycle to repeat. Figure modified from Woolhouse et al., 2002 [5].