Is intron removal by recursive splicing necessitated by the evolution of genomic gigantism?

Splicing of introns is an important step that pre-mRNA transcripts undergo during processing in the nucleus to become mature mRNAs. Although long thought to occur in a single step, introns are now also known to be removed in multiple steps through a process called recursive splicing. This non-canonical form of splicing is hypothesized to increase splicing fidelity, particularly in longer introns. In species that have evolved gigantic genome sizes, overall intron lengths are much longer than in genomes of typical size. Using West African lungfish ( Protopterus annectens ; genome size ∼40Gb) as a model, we use total RNA-seq data to test the hypothesis that gigantic genomes have shifted to rely predominantly on recursive splicing to manage their long introns. Our results suggest levels of recursive splicing at conserved sites in the lungfish that are similar to those seen in humans, indicating that genome-wide intronic expansion accompanying genomic gigantism does not necessitate ubiquitous use of recursive splicing. However, in addition to these results, we also observed patterns of decreasing RNA-seq read depths across entire intron lengths and note that both canonical co-transcriptional splicing and stochastic recursive splicing using many random splice sites could produce this pattern. Thus, we infer canonical co-transcriptional splicing and/or stochastic recursive splicing – but not ubiquitous recursive splicing at conserved sites – manage the removal of long introns in gigantic genomes. We conclude that the vast increase in intron length in gigantic genomes does not appear to have coincided with a fundamental shift in splicing mechanisms.