Primary sequence context defines recursive splicing in long genes

Abstract

Recursive splicing (RS) is a splicing mechanism to remove long introns from messenger RNA precursors of long genes. Compared to the hundreds of RS events identified in humans and drosophila, only ten RS events have been reported in mice. To further investigate RS in mice, we analyzed RS in the mouse brain, a tissue that is enriched in the expression of long genes. We found that nuclear total RNA sequencing is an efficient approach to investigate RS events. We analyzed 1.15 billion uniquely mapped reads from the nuclear total RNA sequencing data in the mouse cerebral cortex. Unexpectedly, we only identified 20 RS sites, suggesting that RS is a rare event in the mouse brain. We also identified that RS is constitutive between excitatory and inhibitory neurons and between sexes in the mouse cerebral cortex. In addition, we found that the primary sequence context is associated with RS splicing intermediates and distinguishes RS AGGT site from non-RS AGGT sites, indicating the importance of the primary sequence context in RS sites. Moreover, we discovered that cryptic exons may use an RS-like mechanism for splicing. Overall, our findings provide novel insights into RS in long genes.