Understanding splicing regulation through RNA splicing maps

Abstract

Alternative splicing is a highly regulated process that greatly increases the proteome diversity and plays an important role in cellular differentiation and disease. Interactions between RNA-binding proteins (RBPs) and pre-mRNA are the principle regulator of splicing decisions. Findings from recent genome-wide studies of protein-RNA interactions have been combined with assays of the global effects of RBPs on splicing to create RNA splicing maps. These maps integrate information from all pre-mRNAs regulated by single RBPs to identify the global positioning principles guiding splicing regulation. Recent studies using this approach have identified a set of positional principles that are shared between diverse RBPs. Here, we discuss how insights from RNA splicing maps of different RBPs inform the mechanistic models of splicing regulation.

Figure 1

Schematic procedure for generating an RNA splicing map. (a) RNA splicing maps are generated by integrating RBP-binding data and splicing profiles. A variety of methodologies can provide this information depending on the model system, technology available and research goals. The simplest form of an RNA splicing map summarises the effects of RBP (ellipses) binding on a cassette exon. Exons are separated into those enhanced (red line) or silenced by the RBP (blue line), and control exons, which are not regulated by the RBP. To find positional principles of splicing regulation, RBP-binding data are combined for each group of exons into a hypothetical, composite cassette exon, with a focus on positions surrounding (typically within 200 base pairs) the 5′ and 3′ splice sites (5′ SS and 3′ SS), the potential branch points of the alternative exon (green box) and the flanking exons (grey boxes). (b) A Nova RNA splicing map for cassette exons generated by integrating the bioinformatic identification of Nova-binding sites and splice-junction microarray data (reproduced with permission from [20]). (c) A Nova RNA splicing map for cassette exons generated by integrating the HITS-CLIP experimental identification of Nova-binding sites and splice-junction microarray data (reproduced with permission from [32]). (d) An RNA splicing map for Pasilla, the Drosophila orthologue of Nova, generated by integrating the bioinformatic identification of Pasilla-binding sites and splicing profiles from RNA-seq data (reproduced with permission from [34]). As shown for the Nova RNA splicing maps in (b) and (c), Pasilla-binding sites are most enriched within and immediately upstream of the skipped exons (blue line) and downstream of enhanced exons. The distance relative to the closest splice site is shown underneath each map. The position of RBP binding is shown on the x-axis. The frequency of RBP binding is shown on the y-axis in red for enhanced, blue for silenced and black for control exons.

Figure 2

Shared positional effects identified by the RNA splicing maps of different RBPs. (a) Positions in silenced exons with the enriched binding of different RBPs. Shown above the transcript diagram, RBP (blue ellipse) binding to a single site close to 3′ SS silences exon inclusion (blue line). Even though not shown in this diagram, such a site can also lie within the exon or close to the 5′ SS. Shown below the transcript diagram, some RBPs bind at intronic positions close to 3′ and 5′ splice sites of the exon to efficiently silence exon inclusion. The arrows indicate that binding at the different positions is achieved by the different RBPs or a hnRNP C. (b) Positions of enhanced exons with the enriched binding of different RBPs. Shown above the transcript diagram, RBP (red ellipse) binding downstream of the 5′ splice site of a cassette exon promotes its inclusion (red line). Shown below the transcript diagram, RBP binding at multiple positions at both ends of the downstream intron enhances exon inclusion. The arrows indicate that interaction between the RBPs bound at both sides of the intron might be required for the enhancing effect. In contrast to the RBPs studied so far, SR proteins enhance inclusion when bound within exons. Upstream and downstream exons (grey boxes) and potential branch points (yellow boxes) are also indicated for positional reference.

Figure 3

The distal regulation of variable-length exons by TIA proteins. (a) TIA protein (red ellipse) binding downstream of the 5′ splice site (5′ SS) stabilises U1 snRNP (yellow ellipse) binding, thereby potentially increasing the rate of transition from the H complex (RBP assembly) to A complex (splice site pairing), which we refer to as faster splicing kinetics. This reduces the time available for RBPs, such as SR proteins (black ellipse), to assemble on the nascent transcript, leading to the preferential exclusion (thick blue line) of the variable region (green box) relative to inclusion (thin red line). Local enhancement leads to distal silencing. (b) In the absence of TIA proteins, the reduced stability of U1 snRNP binding might slow splicing kinetics and thereby increase the temporal window for the assembly of RBPs on the nascent transcript. For example, the binding of additional SR proteins (marked by a question mark) could promote the use of an alternative 3′ splice site (3′ SS), resulting in the preferential inclusion (thick red line) of the variable region of the exon. The blue line represents exon skipping, red line represents exon inclusion and the thickness of the lines represents the extent of the splicing choice.

Figure 4

Asymmetric splicing decision making. Splicing decisions for cassette exons (green box) are the result of the exon skipping (blue line) pathway competing with two exon inclusion (red lines) pathways. RNA splicing maps indicate that RBPs often enhance splicing by binding to the downstream intron. As a result, these RBPs asymmetrically enhance the downstream intron removal pathway of exon inclusion. In this schematic example, RBP (red ellipse) assembly on the nascent transcript promotes the pathway where the downstream intron is spliced first. (a) In the absence of sufficient RBP activity, perhaps because of fast splicing kinetics reducing the temporal window for RBP assembly, the exon skipping pathway dominates the competition. (b) Slower splicing kinetics might allow the increased binding of SR proteins (black ellipse) to the exon, and other RBPs downstream of the exon that promote stable U1 snRNP (yellow ellipse) binding to the 5′ splice site (5′ SS). In this case, increased RBP assembly enhances exon inclusion by promoting the removal of the downstream intron (red line).