Landscape of retron diversity across the SPIRE prokaryotic metagenome resource reveals candidate novel type XI-like lineages

Retrons are prokaryotic genetic elements encoding a specialized reverse transcriptase (RT) that synthesizes multicopy single-stranded DNA (msDNA) and are increasingly recognized as components of bacterial anti-phage defense systems. However, their diversity and ecological distribution across large-scale genomic resources remain poorly characterized. Here, we surveyed retron RTs across the SPIRE representative metagenome collection, a non-redundant, species-level dataset spanning diverse microbial habitats. Using type-specific hidden Markov models cross-calibrated against all known retron groups, we identified retrons representing all canonical types together with additional divergent lineages. Retron distribution was highly structured across both taxonomy and ecology, with some lineages restricted to specific bacterial phyla and others broadly distributed across habitats. Host-associated environments accounted for the largest fraction of retron candidates, although several retron types showed clear specialization for aquatic or anthropogenic niches. Systematic novelty assessment identified two candidate type XI-like lineages, TXI_C2like and TXI_noncan_h, characterized by protease-independent architectures and lineage-specific accessory modules. De novo computational analyses further identified candidate msr/msd-like non-coding RNA elements upstream of the RT genes in both lineages, indicating that these highly divergent systems retain the defining RT-ncRNA architecture of canonical retrons. Together, these findings expand the known diversity of retron systems and reveal a strong interplay between evolutionary history and ecological specialization.