Drosophila Down Syndrome Cell Adhesion Molecule: A paradigm for revealing hidden splicing codes

The exciting prospect of exploiting genome information for personalized medicine critically depends on the extent to which we understand the regulatory information residing outside the protein-coding regions of the genome. A unique feature of genes in eukaryotic organisms is their organization into protein-coding DNA sequences, termed exons, which are separated by non-coding introns.

During splicing, introns are excised from the pre-messenger RNA (mRNA) transcript by the spliceosome and exons are joined to form the mature mRNA. A functional protein can then be made from the mRNA, but only if splicing controlled by hundreds of proteins has accurately taken place. The unique organization of eukaryotic "genes in pieces" further allows exons to be included in one mRNA from a particular gene, but excluded in another.

This process, termed alternative splicing (AS), is used in most human genes and is an important mechanism to build complex organisms with comparatively few genes. AS is particularly prevalent in the brain and changes during aging. Mis-regulation of AS is also associated with various human diseases, including cancer, metabolic disorders and neurodegeneration.

Fidelity of splicing rests critically on accurate reading of 'splicing information' in non-coding regions of the pre-mRNA. Paradoxically, introns are often very large and contain numerous sequence motifs that look like splice sites. Hence, the splicing information is encrypted in a code of short sequence motifs that we do not understand very well.

As the splicing process is very complex, it is also vulnerable to cause human disease from mutations present in our genomes that result in aberrant splicing. In fact, about 15% of genetic human disease is caused by mutations in splice sites, but considering all regulatory elements involved in splicing, estimates range up to 50%.

However, a drug consisting of a short stretch of nucleotides has recently been approved in the US and the EU for correction of splicing in the Spinal Muscular Atrophy (SMA) gene. Since such drugs can be directed to any part in the genome, many cases of aberrant splicing causing human disease could potentially be corrected. To make full use of this technology we need to understand the splicing code.

The fruit fly Drosophila has proven an excellent and cost-effective genetic model for deducing basic biological processes as illustrated by the 2017 Nobel prize award. To discover fundamental splicing codes the Down Syndrome Cell Adhesion Molecule (Dscam) is an excellent model gene, because it is extensively alternatively spliced in four arrays of variable exons where only one exon is chosen for inclusion in the mature mRNA.

This way, 36'016 different protein isoforms can be generated, which are more proteins from one gene than genes are present in the genome. This diversity is essential for development of the brain, but also in the immune system for recognition and clearance of pathogens such as bacteria.

The most central questions regarding Dscam AS is why exons in the variable clusters are not spliced together despite having consensus splice sites and how one variable exon is chosen. We now have developed a toolkit in the fruit fly Drosophila that allows us to test a) whether splicing together of variable exons is prevented by splicing signals being too close together to allow for assembly of a functional splicesome, b) whether long-range base-pairings are key to Dscam AS to bring a variable exon into the proximity of flanking constant exons and c) whether each variable cluster contains unique regulatory sequences that restrict splicing to specific parts of a gene.

From these experiments we will learn about fundamental mechanism involved in AS regulation and how their mis-regulation can lead to human disease. Our results will be instrumental for elucidating the splicing code to instruct how human disease caused by splicing errors can be corrected.