Lack of evidence for gene-level convergence linked to evolutionary shifts in torpor among placental mammals

Abstract

Torpor is a key survival strategy that many avian and mammalian lineages evolved in response to challenging environmental conditions. Whether the independent evolution of torpor in different lineages involved changes in the same genes remains poorly understood. Here, we performed comparative screens across 190 placental mammal genomes to comprehensively examine associations between loss, positive selection and evolutionary rate shifts in individual protein-coding genes and evolutionary shifts in torpor use. We find that gene-torpor associations are highly clade-specific, with no gene being able to explain the majority of torpor shifts across the phylogeny of placental mammals. Instead, a relatively higher but limited extent of evolutionary convergence can be detected at the pathway level. Our results suggest that torpor emerged through several genetic routes in placental mammals, which likely explains the vast diversity of torpor use patterns that can be observed among torpor-capable species today. SignificanceA fundamental question in evolutionary genomics is whether independent gains and losses of a convergent trait may be achieved through similar or distinct genetic paths. Here, we address this question by focusing on the evolution of torpor across placental mammals, a key trait that likely emerged independently in several lineages. By conducting multiple comparative genomics screens across 190 placental mammal species, we find that individual protein-coding genes have a low explanatory power for evolutionary shifts in torpor. In contrast, there is slightly stronger evidence for pathway-level convergence. Our findings suggest that the wide diversity of patterns of torpor use across placental mammals may, in part, be due to the existence of multiple genetic paths to torpor.