Coevolution between hosts and symbionts is fundamentally asymmetric. Symbiotic mutualists or parasites can adapt to their hosts faster than hosts can adapt in response because the symbionts usually have shorter generation times — and they also generally have the benefit of much larger population sizes, providing a bigger pool of potentially useful mutations and reduced influence of genetic drift. This latter advantage, though, can be lost to the ecological consequences of coevolution with hosts. If hosts evolve resistance to a parasite, the parasite population will collapse, until a new counter-resistance mutation emerges.
How this kind of ecological feedback affects the coevolution of hosts and symbionts is a challenging thing to track, but a paper published recently in Science Advances manages to do it with a genomics-enabled experimental model of host-parasite coevolution. Tracking the population sizes and genomic diversity of the unicellular alga Chlorella variabilis and a DNA virus that attacks it, the authors identify how host and parasite population dynamics shape host-parasite coevolution.
The study’s coauthors, led by Cas Retel at EAWAG in Switzerland, coevolved three replicate populations of Chlorella variabilis and the virus in chemostats for 100 days, tracking host and virus densities on a daily basis by flow cytometry, and performing paired-end Illumina sequencing of samples taken at 12 timepoints. All of that data comes together beautifully in their fourth figure, which links selective sweeps — rapid changes in the frequency of individual host or virus alleles — with changes in host and virus population size.
The Chlorella genome is, as you might expect, a good bit more complex than the viral genome — where the host genome is 46 million basepairs long, with only about 26% devoted to protein-coding genes, the viral genome is 300 thousand basepairs long, with 79% gene content. In the host, over half of loci showing substantial changes in allele frequency were synonymous sites or in noncoding regions, which the authors interpret as evidence of genetic hitchhiking; the overwhelming majority of allele frequency changes in the virus were at nonsynonymous sites, mostly in three genes with known roles in infectivity. That is, most of the observed evolution by the virus was likely due to direct selection.
The diversity of the two species changed in concert with changes in population size and selective sweeps. In periods immediately following sweeps in the hosts, the data show both population expansion and a distribution of allele frequencies consistent with population growth under the neutral model. This is a bit of a mind-bender, but it’s an outcome that makes sense: Right after a sweep, the population is fixed for a mutation that confers resistance to the virus, so it’s free to grow. Precisely because the whole population carries that new resistance allele, post-sweep growth is selectively neutral — and this allows for the accumulation of new variation after the sweep has eliminated it.
This is, in micro-scale, ecological release, in which "escape" from an antagonistic interaction frees a population to expand and explore new variation. The timing of this dynamic, Retel et al. report, is about as rapid as the loss of diversity during the course of a selective sweep. That is to say, the ecological consequence of resistance evolution (population growth) feeds back to provide new raw material for ongoing adaptive evolution.
Retel C, V Kowallik, W Huang, B Werner, S Künzel, L Becks, and PGD Feulner. 2019. The feedback between selection and demography shapes genomic diversity during coevolution. Science Advances 5 (10), eaax0530. doi: 10.1126/sciadv.aax0530