Changing Fitness Landscapes During Host-Parasite Coevolution Opens Adaptive Pathways to Evolutionary Novelty

It has been proposed that the evolution of new functions can be explained by host-parasite coevolution. Experiments on a virus, bacteriophage lambda, and its host, Escherichia coli, have accordingly shown that lambda evolves to exploit a novel receptor during their coevolution. The importance of coevolution for the innovation, however, was not directly tested. We provide a comprehensive analysis of the process of adaptation leading to evolutionary novelty in this model system by measuring lambda’s fitness landscape with a new high throughput technology (MAGE-Seq). The innovation in lambda was thought to be triggered by the evolution of host resistance, which was expected to deform lambda’s landscape in ways to open adaptive pathways. As predicted, the landscape measured with wild type cells was significantly different than when measured with evolved resistant host cells. However, this initial hypothesis turned out to be overly simple. Simulations of lambda populations evolving on the landscapes, along with additional experiments, showed that paths to the innovation required evolving first on the wild type host, and then on the evolved host. Altogether, we provide definitive evidence that coevolutionary dynamics can lead to evolutionary innovations.