Tissue-specific splicing of disordered segments that embed binding motifs rewires protein interaction networks

Abstract

Alternative inclusion of exons increases the functional diversity of proteins. Among alternatively spliced exons, tissue-specific exons play a critical role in maintaining tissue identity. This raises the question of how tissue-specific protein-coding exons influence protein function. Here we investigate the structural, functional, interaction, and evolutionary properties of constitutive, tissue-specific, and other alternative exons in human. We find that tissue-specific protein segments often contain disordered regions, are enriched in posttranslational modification sites, and frequently embed conserved binding motifs. Furthermore, genes containing tissue-specific exons tend to occupy central positions in interaction networks and display distinct interaction partners in the respective tissues, and are enriched in signaling, development, and disease genes. Based on these findings, we propose that tissue-specific inclusion of disordered segments that contain binding motifs rewires interaction networks and signaling pathways. In this way, tissue-specific splicing may contribute to functional versatility of proteins and increases the diversity of interaction networks across tissues.

Graphical abstract

Figure 1

Classification of Exons and Characterization of the Protein Segments (A) Cassette exons that are tissue-specifically spliced are in red, and alternatively spliced cassette exons not present in the set of tissue-specific exons are in dark gray. Constitutive exons are in light gray. (B) Shown are types of data and computational approaches used to characterize the exons. See also Figure S1.

Figure 2

Tissue-Specific Segments Are Enriched in Disordered Binding Motifs and Are Conserved Percentages of protein segments encoded by the three different sets of exons that contain (A) Pfam domains, (B) disordered regions, (C) ANCHOR-binding motifs, and (D) PTM sites. Significance was assessed by chi-square test, and p values are indicated in each panel. Box plot of the distribution of conservation (percent amino acid identity) for the (E) disordered region, (F) binding motifs, and (G) binding motifs versus other regions for the three different sets of exons. Statistical significance was assessed using Mann-Whitney test. The median value within each set is shown with a thick black line. Boxes enclose values between the first and third quartile. Interquartile range (IQR) is calculated by subtracting the first quartile from the third quartile. All values that lie more than 1.5× IQR lower than the first quartile or 1.5× higher than the third quartile are considered to be outliers and were removed from the graphs to improve visualization. The smallest and highest values that are not outliers are connected with the dashed line. The notches correspond to ∼95% confidence interval for the median. See also Figure S2 and Table S1.

Figure 3

Genes with Tissue-Specific Exons Play an Important Role in Protein Interaction Networks (A) Distribution of the number of interaction partners for genes with a TS exon (TSE genes; with and without predicted binding motif) and for all other genes (non-TSE genes). See Figure 2 legend for the description of a box plot. (B) Average number of interaction partners for the TSE (red circles) and non-TSE (gray circles) genes in different tissues. The average and median number of interactions for TSE genes was greater than for non-TSE genes in each of the tissues (p < 5 × 10⁻²; Mann-Whitney), except for liver and testis interstitial cells. (C) Principle of the Jaccard similarity index. The score for a gene varies between 0 and 1, where 0 indicates that no interactions are maintained and 1 indicates that all interactions are maintained between a pair of tissues. Hence, the smaller the number is, the more tissue specific are its interactions. (D) Distribution of Jaccard similarity indices for TSE and non-TSE genes in the tissues included in this study. Gray horizontal line in all box plots indicates median value for the non-TSE genes. Statistical significance was assessed using Mann-Whitney test. See also Figure S3 and Table S2.

Figure 4

Examples of Tissue-Specific Exons that Can Affect Protein Interactions Transcripts (above) and the encoded protein sequences (below) are shown for tissues (box). Protein domains (colored regions) and tissue-specific segments (red) are shown. (A) The PIP5K1C gene shows differential inclusion of a TS exon between cerebellum and lymph node. This segment undergoes disorder-to-order transition when interacting with the clathrin adaptor AP-2 (AP2B1). The exclusion of this segment is likely to abolish the interaction. (B) Protein structure (3H1Z), containing interacting segments from these two proteins. (C) PACSIN2 shows differential inclusion of a TS exon between cerebellum and breast tissue. All four annotated phosphorylation sites are inside the TS exon. Thus, the exclusion is likely to affect regulation via phosphorylation. See also Table S3.

Figure 5

Alternative Inclusion of Tissue-Specific Exons Can Rewire Interaction Networks and Modulate Protein Interactions (A) When tissue-specific splicing gives rise to isoforms that differ in the presence of interacting protein segments, this can result in the rewiring of an interaction network in the respective tissues. This is illustrated with the PIP5K1C kinase gene that has an exon with different inclusion levels in cerebellum and lymph mode. The exon encodes a binding motif (red cartoon representation), which mediates the interaction with the AP-2β appendage domain (cyan surface representation). (B) Molecular mechanisms by which a tissue-specific segment can rewire or fine-tune protein interactions. Segments in red are encoded by TS exons; other regions encoded by the same protein are colored dark gray. Interaction partners are colored in cyan. Segments encoded by TS exons can include a region that directly interacts with other proteins, DNA, RNA, or ligands, or indirectly affect the protein's binding properties (e.g., affinity, kinetics, and selectivity). TS segments that are involved in interactions are less frequently domains and more often disordered regions that embed peptide-binding motifs. An example for the former is a TS segment encoding the complete PF21A PHD zinc finger domain that recognizes nonmethylated lysine 4 of histone H3 (2PUY). An example for the latter is the disordered segment encoded by a TS exon of PEX19, which becomes ordered upon binding to PEX3 (3MK4). An example of a splicing event that affects linker length, thereby altering its affinity to bind DNA, is the fly UBX gene (1B8I; disordered linker, red line; and DNA, light gray surface). An example of where the spliced exon encodes an intramolecular interaction motif that competes or masks an interaction interface is the acid box region of the FGFR kinases (model based on 2CKN and 1YRY; acidic linker, red line). See also Table S4.