Evolutionary reconstruction of bat influenza A viruses reveals segment-specific histories and host-associated selection

Abstract

Influenza A (IAV) is a major zoonotic pathogen infecting a broad range of vertebrate hosts. In this study, we focus on inferring the evolutionary history of a subset of novel IAV, isolated from New World bats, which use a non-canonical host receptor and infer their relationship to sialic-acid-binding IAV. Little is known about when these novel subtypes emerged and how selection has acted upon them compared to sialic-acid-binding circulating strains. The current state of the field is split on whether bats are natural reservoirs for a distinct yet low abundance IAV lineage, or if influenza underwent rapid evolution after a recent spillover from a species within its standard host range. By analyzing segment-specific phylogenies, we hope to clarify the evolutionary history of bat IAV. Our objectives are to infer the date of divergence of non-canonical bat IAV from sialic-acid binding IAV, identify patterns of selection acting on bat IAV compared to representative samples of sialic-acid-binding IAV to identify potential signatures of genome-wide adaptation, and finally to reconstruct ancestral state transitions to infer which host taxa preceded spillover into bats. To accomplish this, we analyzed each IAV genome segment independently using least-squares molecular dating, Bayesian discrete-trait ancestral host reconstruction, and branch-site selection tests. Our results suggest that non-canonical bat IAV diverged from major IAV lineages 100-200 years ago, supporting a deeper evolutionary history rather than a very recent spillover event. However, discordance among genome segments in inferred divergence times, selective pressures, and ancestral hosts indicates that bat IAV experienced segment-specific evolutionary histories. Together, these findings suggest that non-canonical bat IAV represents a deeply divergent lineage with a complex evolutionary history involving segment-specific selection and host-transition dynamics.