Alternative splicing and its potential effects on protein structure. During transcription, RNA is generated from the chromosome’s genetic material. This piece of RNA forms the basis for the mRNA that will be translated into proteins, hence the designation “pre-mRNA.” mRNA is composed of terminal UTRs, which help guide and regulate the translation process, and a central coding region, which is translated into the corresponding protein product. Eukaryotic genes contain fragments that are not used in the translation process, called introns, and regions that will become part of the final mRNA, called exons. Here exons are shown as colored cylinders. The mRNA to be translated is assembled from pre-mRNA by splicing the introns away, so that only the exons remain, as shown in both sides of the image. During alternative splicing, some exons also can be partially or completely spliced away, as can be seen in the path marked by purple arrows to the right of the figure, where the red exon (and two of the UTR exons) are “skipped,” generating a shorter mRNA and hence a shorter protein product. The different products of alternative splicing are called isoforms. In this work, the region missing in isoform 2 because of alternative splicing is called an AS region. The generation of isoforms by alternative splicing poses potential problems for the structural stability of globular proteins, because protein folding is a strongly cooperative process. Indeed, the absence of the red AS region in isoform 2 can potentially disturb the protein’s 3D structure very strongly, but few systematic studies of this effect have been carried out. (Note: Although we have described this process in terms of removal of protein regions, the same argument can be made about insertion of AS regions, without loss of generality.)

Distribution of 75 AS regions according to the incidence of disorder. Regions of proteins encoded by sections of pre-mRNA that are spliced out in different protein isoforms were compared with regard to their structural characterization. The largest proportion of the protein segments (43 of 75, or 57%) is entirely disordered. These fragments encompass a total of 4,929 aa residues and range in length from 1 to 1,183. The 18 partially disordered segments (24%) range in length from 8 to 570 and include a total of 820 disordered and 1,581 ordered residues. Only 14 segments, or 19%, correspond to fully ordered regions. These segments range in length from 1 to 147 and comprise 613 aa residues. To obtain these results, a set of isoform transcripts corresponding to each protein in the ASED data set was obtained from the ASG database. Regions that are spliced out from pre-mRNA were found by translating those isoform transcripts and aligning the resulting protein sequences to that of the original ASED protein. Any segment from the original ASED protein that is missing from an isoform sequence is considered a spliced-out region. Only spliced-out regions that were fully characterized structurally were used to construct this graph.

Disorder content distributions of the studied data sets and their corresponding regions affected by alternative splicing. For proteins or regions in the ASED data set, the plots show bars for both experimentally determined disorder and predicted disorder. For the ASSP data set, the chart shows the predicted disorder content distribution. (a) Histogram of disorder content per protein. (b) Histogram of disorder content per region affected by alternative splicing. The error bars represent a 68% confidence interval or 1SD.

Functional domains of BRCA1 and representative isoforms. The red line marked with the horizontal arrows on top indicates the extent of the 1,500-aa-long disordered central region. The ordered RING and BRCT domains are also shown at the termini. The vertical arrows show the location of phosphorylation sites, and the colored rectangles represent binding domains, including the binding partner name (see text). Isoforms are shown at the bottom, with dotted lines representing the regions missing from the translated protein products due to alternative splicing. The numbers at the left are the ASG isoform identifiers. The missing numbers (2, 3, 7, and 9) refer to isoforms that involve insertions or additions for which we have no structural information.