Microevolutionary cophylogeny reflects host-symbiont population dynamics and human mitonuclear interactions

Abstract

Cophylogeny, the study of phylogenetic similarity between interacting organisms, provides insights into the specificity and shared evolutionary history of symbiosis. While the ecological drivers of cophylogeny have been investigated at the macroevolutionary scale, the influence of these processes on microevolution remains unclear. This is due, in part, to the fact that the ancestral relations between individuals within a sexually reproducing eukaryotic host species cannot be well represented with a single phylogenetic tree, since genetic distances between individuals change substantially across the genome due to meiotic recombination. This heterogeneity can be captured and utilized through the inference of an ancestral recombination graph (ARG) built from the genomic data of the host. Here, we propose to measure microevolutionary cophylogeny by comparing a symbiont evolutionary tree to a host ARG. This approach simultaneously measures genome-wide cophylogeny, as well as locus-specific signals. Through simulations, we investigate the effects of transmission mode, population structure, admixture, and allelic incompatibility on microevolutionary cophylogeny. In contrast to macroevolutionary patterns, we find a limited relationship between cophylogeny and vertical transmission, with vertically transmitted host-symbiont systems displaying no cophylogeny in large panmictic populations. We apply our approach to mitochondrial and nuclear genomes within the 1000 Genomes Project--a host-symbiont system with strict maternal transmission--and observe substantial variation in mitochondrial-nuclear (mitonuclear) cophylogeny across human populations. Finally, we investigate locus-specific signals of cophylogeny and observe limited evidence of mitonuclear incompatibility.