Quantifying the strength of viral fitness trade-offs between hosts: a meta-analysis of pleiotropic fitness effects

The range of hosts a given virus can infect is widely presumed to be limited by fitness trade-offs between alternative hosts. These fitness trade-offs may arise naturally due to antagonistic pleiotropy if mutations that increase fitness in one host tend to decrease fitness in alternate hosts. Yet there is also growing recognition that positive pleiotropy may be more common than previously appreciated. With positive pleiotropy, mutations have concordant fitness effects such that a beneficial mutation can simultaneously increase fitness in different hosts, providing a genetic mechanism by which selection can overcome fitness trade-offs. How readily evolution can overcome fitness trade-offs therefore depends on the overall distribution of mutational fitness effects between hosts, including the relative frequency of antagonistic versus positive pleiotropy. We therefore conducted a systematic meta-analysis of the pleiotropic fitness effects of viral mutations reported in different hosts. Our analysis indicates that while both antagonistic and positive pleiotropy are common, fitness effects are overall positively correlated between hosts and unconditionally beneficial mutations are not uncommon. Moreover, the relative frequency of antagonistic versus positive pleiotropy may simply reflect the underlying frequency of beneficial and deleterious mutations in individual hosts. Given a mutation is beneficial in one host, the probability that it is deleterious in another host is roughly equal to the probability that any mutation is deleterious, suggesting there is no natural tendency toward antagonistic pleiotropy. The widespread prevalence of positive pleiotropy suggests that many fitness trade-offs may be readily overcome by evolution given the right selection pressures.