A large-scale phylogeny of replication initiation proteins illuminates plasmid macroevolutionary landscape

SUMMARY Plasmids are the most influential engines of bacterial evolution and horizontal gene transfer, fueling the global spread of traits such as antimicrobial resistance. Their deep evolutionary relationships, however, remain difficult to resolve because current classification schemes are constrained by host range and nucleotide similarity. Replication initiation proteins (RIPs), which govern plasmid persistence and diversification, also remain poorly annotated across public databases. Here we establish PInc, a curated and experimentally grounded replicon classification framework anchored in historically defined incompatibility groups of Pseudomonas plasmids. Homology searches beyond PInc revealed that most replication initiators analyzed here share a conserved winged-helix (WH) domain, defining a broad WH RIP superfamily. Using the conserved WH region, we reconstructed a large-scale phylogeny that linked WH RIPs to over 100,000 plasmids, representing approximately half of those analyzed across public databases. This phylogeny resolved eight major clades and the deep split between the single- and double-winged-helix superclades, while revealing clade-specific host and environmental distributions and substantial RIP diversity not captured by current typing tools or annotation schemes. Together, these results overcome the historical host bias of plasmid typing and provide a replication-centered view of plasmid diversification across bacterial lineages and environments.