The subordinate role of pseudogenization to recombinative deletion following polyploidization in angiosperms

Extensive gene loss is a hallmark of rediploidization following polyploidization, but its molecular basis remains unclear: whether it occurs primarily through pseudogenization or DNA deletion. Here, we examine pseudogenization in collinear segments from ancient whole-genome multiplications (WGMs) across 12 angiosperms. Although total pseudogenes are abundant, we find far fewer WGM-derived pseudogenes than expected if pseudogenization and DNA deletion contribute equally to gene loss. Simulations of neutrally evolving pseudogenes indicate that, if DNA deletion is absent, pseudogenes should be detectable for far longer than observed in the paleo-polyploid genomes, suggesting gene loss driven by DNA deletion. Analyses of three neo-autopolyploid genomes confirm this pattern: among substantial gene loss, DNA deletions occur on average 1.5 times more frequently than pseudogenization. Our findings imply that gene loss post-polyploidization primarily takes place via DNA deletion, enabled by a genomic environment with an elevated recombination rate created by WGMs. In contrast, small-scale duplications yield scattered duplicated genes, which appear less exposed to deletion and hence result in a high number of pseudogenes. This model is further reinforced by an enrichment of WGM-derived pseudogenes in high recombination regions. Moreover, some pseudogenes may govern a function, as indicated by non-neutral K_a/K_s ratios and overlap with lncRNAs.