Competing RNA secondary structures are required for mutually exclusive splicing of the Dscam exon 6 cluster

Abstract

Alternative splicing of eukaryotic pre-mRNAs is an important mechanism for generating proteome diversity and regulating gene expression. The Drosophila melanogaster Down Syndrome Cell Adhesion Molecule (Dscam) gene is an extreme example of mutually exclusive splicing. Dscam contains 95 alternatively spliced exons that potentially encode 38,016 distinct mRNA and protein isoforms. We previously identified two sets of conserved sequence elements, the docking site and selector sequences in the Dscam exon 6 cluster, which contains 48 mutually exclusive exons. These elements were proposed to engage in competing RNA secondary structures required for mutually exclusive splicing, though this model has not yet been experimentally tested. Here we describe a new system that allowed us to demonstrate that the docking site and selector sequences are indeed required for exon 6 mutually exclusive splicing and that the strength of these RNA structures determines the frequency of exon 6 inclusion. We also show that the function of the docking site has been conserved for ~500 million years of evolution. This work demonstrates that conserved intronic sequences play a functional role in mutually exclusive splicing of the Dscam exon 6 cluster.

FIGURE 1.

The Dscam BAC system. (A) D. melanogaster S2 cells were either untreated (−) or transfected (+) with a BAC containing the Dscam gene (Dscam^5/7vir BAC). RT-PCR was performed using exon 5 and 7 primers for either D. melanogaster or D. virilis. (B) The RT-PCR products were excised and sequenced using the multiplex sequencing method and the percent inclusion of each exon was calculated and represented as a heat map.

FIGURE 2.

Dscam docking site mutations. (A) The Dscam^5/7vir BAC docking site (in red) was either completely deleted (ΔDock), the 5′ side was deleted (ΔLeft), or the 3′ side was deleted (ΔRight). (B) D. melanogaster S2 cells were transfected with Dscam^5/7vir BAC docking site mutant constructs. RT-PCR was performed with primers that anneal to BAC exons 5 and 7 (blue), BAC exon 7 and the junction of exons 7 and 8 (orange), or endogenous exons 5 and 7 (purple). When Dscam exon 6 was amplified, two products were observed, one which includes an exon 6 variant (6.x) and one lacking an exon 6 variant. (C) Dscam exon 6 skipping was measured for each mutant docking site BAC construct. Quantitative RT-PCR was used to measure skipping compared to the unaltered (WT) Dscam^5/7vir BAC. The log₂ fold change compared to WT is plotted for each mutant BAC construct with positive and negative error bars.

FIGURE 3.

The Daphnia pulex Dscam docking site is functionally conserved. (A) D. melanogaster and D. pulex Dscam gene organization. (B) The docking site sequences from D. melanogaster and D. pulex are shown. The conserved sequence is highlighted in orange. (C) D. melanogaster S2 cells were transfected with the D. pulex Dscam minigene (WT) and a minigene in which the docking site was deleted (ΔDock). RT-PCR was performed with primers that anneal to D. pulex exons 5 and 7. Two products were observed, one that includes an exon 6 variant (6.x) and one that lacks an exon 6 variant.

FIGURE 4.

Docking site mutations impact the relative frequency of exon 6 inclusion. The mutant docking site constructs were transfected into S2 cells in triplicate and the RT-PCR products were excised and sequenced using the multiplex sequencing method. (A) The log₂ fold change of each mutant construct compared to WT Dscam^5/7vir BAC is represented as a heat map. (B) Percent inclusion is plotted for each exon 6 variant. (C) The change in the strength of the predicted docking site-selector sequence interactions (ΔΔG) are plotted as a function of the log₂ fold change in the frequency of inclusion for each docking site deletion construct tested.

FIGURE 5.

Selector sequences determine the efficiency of exon 6 inclusion. (A) The selector sequences of individual exons were replaced in the Dscam^5/7vir BAC as indicated on the left. Transfection and multiplex sequencing were used to determine the fold change of exon 6 inclusion with respect to the unmodified BAC. The fold change (log₂) was plotted as a heat map. In each experiment the exon containing the modified selector sequence is highlighted in red. (B) The change in the strength of the predicted docking site-selector sequence interaction (ΔΔG) is plotted as a function of the fold change in the frequency of inclusion of the modified exon 6 variant.