Figure 1. Mechanisms of splicing regulation: (a–b) activation; (c–e) repression. Splicing activation is enabled primarily through the activator function of the RS domain of an SR protein bound to an ESE element through its RRM domain (a), or can be RS-domain independent, with the activator blocking the action of a nearby silencing element in direct competition (b). Splicing repression is accomplished primarily by hnRNP proteins, either by neutralizing the effect of an activator protein through direct competition (c), or by cooperative binding of several silencing factors that displace the ESE-bound activators (d), or by binding to motifs on either side of the exon followed by dimerization, which places the exon in a loop inaccessible to the splicing apparatus (e). Reproduced from Florea (2006).

Figure 2. Types of alternative splicing events: (a) exon inclusion/exclusion (‘skipping’); (b) alternative 3¢ exon end; (c) alternative 5¢ exon end; (d) intron retention; (e) alternative 5¢ UTR and promoter region and (f) alternative 3¢ UTR and polyadenylation site. Exons are represented by boxes, and introns by straight or segmented lines connecting the exons. Alternatively spliced elements (exons or portions of exons) are shown in dark grey, and those constitutively spliced in light colour.

Figure 3. Bioinformatics methods for identifying alternative splicing events: (a) identification of alternative splicing events by direct comparison of cDNA sequences, (b) comparison of exon–intron structures on the genome and (c) microarray experiments. In comparisons between pairs of cDNA sequences (a), alternatively spliced features appear as insertions in one sequence compared to the other. Comparison of exon–intron structures from spliced alignments of cDNAs on the genome (b) reveals the exons and introns undergoing alternative splicing. In (c), microarray GeneChip expression data using multiple probes per gene show potential alternative splicing events as groups of probes that are differentially expressed between the two experiments (probes p4 and p5 in exon 2). Reproduced from Florea (2006).

Figure 4. Classification and evolutionary characterization of alternative splicing events. (a) Categories of splicing events: constitutive, alternative minor-form and alternative major-form exons. (b) Determining the time of exon creation by comparative genomics. The exon's present (P)/absent (A) vector in the eight species (human, chimp, mouse, rat, dog, chicken, fugu and zebrafish) is used to determine the ancestral form, and the likely time of insertion of the exon into the gene. For instance, the vector (PPAAPAAA) indicates that the exon is present in human, chimp and dog and absent from rodents, chicken and fishes, and therefore is likely to have been inserted into the mammalian lineage after the chicken split, whereas an exon with the vector (PPAAPAAP) is found as far as in fishes, and therefore is (likely) ancestral. (c) Minor-form exons are predominantly recently created (insertions in human, primates, rodents or dog, i.e. the H, C, R, D categories; 85%), unlike major-form exons (20%) and constitutive exons (26%). Bars show the fraction of exons in each category created during the evolution of the eight species, with lighter shades representing recent insertion events.

Figure 5. Conservation analysis of alternative splicing in human and mouse. (a) Types of exon skipping alternative splicing identified and analysed by comparisons of human and mouse mRNA and EST sequences. An exon can be detected as conserved and alternatively spliced in both species (CAS), conserved and alternatively spliced in one species whereas constitutively spliced in the other (CnAS), or present and alternatively spliced in only one species, i.e. genome-specific (GAS). (b) The percentage of total alternatively spliced skipped exons for each category in (a) that were detected using either mRNA or EST evidence from all sources, including tumour and cell lines (left), versus from ‘normal’ sources only (right). Modified with permission from Pan et al. (2005).