Figure 1.
Basic Steps of Pre-mRNA Splicing and Patterns of Alternative Exons
(A) Five small nuclear ribonucleoproteins (snRNPs U1, U2, U4, U5, and U6), an auxiliary splicing factor (U2AF), and many other factors (not represented) organized in the human spliceosome execute the excision of introns. After the recognition of 5′ss, 3′ss, and branch point, respectively, by U1, U2AF, and U2, the intron is first cleaved at the 5′ss and subsequently at the 3′ss, mediated by U4/U6 and U5 (U1, U2, and U4 are detached later during the cycle). The intron remains in the nucleus and is degraded, while ligated exons are transported outside to the cytoplasm.
(B) AS events can be inferred by spliced alignments of mRNAs to genomic DNA (cf. Figure 3), indicated by dashed lines (AS part of exon colored in black), and commonly distinguished in terms of whether mRNA isoforms differ by skipping of an exon (SE), or whether isoforms differ in the usage of a 5′ss or 3′ss, producing an A5E or A3E, respectively. A fourth type, termed retention-type intron, occurs when two isoforms differ by the presence of an unspliced intron in one transcript that is spliced in the other.
(C) More complex types of AS forms can be constructed from canonical splice variants; different isoforms can also be the result of variations at the 5′- and 3′-terminus of transcripts, which are not necessarily due to AS.
Figure 2.
Selection of Splice Patterns of Known Alternative Exons
Selection of splice patterns of known alternative exons of Tra, Sxl, and Dscam genes in D. melanogaster [3,4], and α-Actinin, α-Tropomyosin, Troponin-T, and PTB in H. sapiens [71]. Exon skipping is the predominant AS event in many metazoans and, e.g., has been shown to be involved in tissue- and developmental stage–specific regulation, as well as autoregulation (PTB) [72]. AS products of pre-mRNAs expressed from Tra and Sxl genes are involved in the pathway of somatic sex fate in D. melanogaster, which is regulated by altogether five AS genes at the top of the determination cascade [4]. The “master gene” Sxl is expressed in female flies, where it acts as a negative regulator of splicing. AS of the Dscam gene is known for its theoretically large number of possible different AS products (∼38,000 against ∼14,000 D. melanogaster protein-coding genes), which are derived from four clusters of skipped exons. The regulation of one cluster includes so-called selector-docking sites, which are inverse complementary overlapping sites located in the most 5′-end intron (docking) and upstream of each skipped exon (selector) of this cluster, respectively [73].
Figure 3.
From Sequences to Patterns and Functional Elements
(A) AS events can be computationally inferred by spliced-sequence alignments of complete or partial mRNAs to genomic DNA. A selection of available algorithms and software is listed in Table S1. The sketch shows seven mRNAs with indicated exon junctions (for visual guidance only), the primary transcript structures of which are to be inferred from alignments to genomic DNA (the order of the mRNAs above and below the genomic DNA is the same). In the example shown, the set of mRNAs aligns to five exons (E1 to E5), and the data are consistent with two AS events: E2 alternative 3′ss splicing, and E3 skipping (skipped in the fourth mRNA from the top).
(B) Splicing-regulatory elements are distinguished depending on their location (exon or intron) and their mode of action (enhancing or silencing): 1) exonic splicing enhancer (ESE) elements; 2) exonic splicing silencer (ESS) elements; 3) intronic splicing enhancer (ISE) elements; and 4) intronic splicing silencer (ISS) elements. One can subclassify these elements whether they carry protein-coding information, act in the context of 5′ss and/or 3′ss, or are sequence-conserved across species (indicated by the presence of vertical colored bars).
(C) Often, ESE, ESS, ISE, and ISS elements do not act independently of their sequence context, but can assume antagonistic functions (enhancing versus silencing) in splicing. The color-coded example sequence elements are taken from the literature [27,28,32,38].
Figure 4.
Conservation of AS across Species
(A) Splice patterns of exons of pairs of orthologous genes can be classified into four “pattern conservation” categories, demonstrated here for SE events in H. sapiens/M. musculus: both exons are constitutively spliced (Sh,m)—known as “constitutive conserved exons” (CCEs); the exon of the human gene is alternatively spliced, the mouse one constitutively (SH,m); the exon of the mouse gene is alternatively spliced, the human one constitutively (Sh,M); both exons are alternatively spliced (SH,M)—known as “alternative conserved exons” (ACEs). In addition, one can define two “gain/loss” categories as: the exon of the human gene is alternatively spliced, the mouse exon is absent (SH, m); the exon of the mouse gene is alternatively spliced, the human exon is absent (S h,M).
(B) AS events can be successfully predicted ab initio, that is, from genomic sequence alone. The figure refers to SEs in particular, but results for other classes of AS events have also been described. Published approaches address one or both of two problems: exon classification and exon discovery. Constitutive and alternative exons show different characteristics, which can be used as features for nontranscript methods. Example features include length of the AS exon and surrounding intron; strength of the splice sites; the level of conservation in the exon and surrounding introns; and coding-typic conservation patterns in exon sequences. In addition, some approaches utilize the occurrence of specific sequence features corresponding to splicing regulatory elements (parts of the figure adapted from [50]).