The sequencing of various eukaryotic genomes has demonstrated that a surprisingly small number of genes generate a complex proteome. For example, the estimated 20,000-25,000 human protein-coding genes give rise to 100,000-150,000 mRNA variants as estimated by EST comparison. Array analysis shows that 74% of all human genes are alternatively spliced (Johnson et al. 2003) and a detailed array-based analysis of chromosome 22 and 21 suggests that every protein-coding gene could undergo alternative splicing (Kampa et al. 2004). Extreme examples illustrate the potential of alternative splicing: the human neurexin 3 gene could form 1,728 transcripts (Missler and Sudhof 1998) and the Drosophila DSCAM gene could give rise to 38,016 isoforms, which is larger than the number of genes in Drosophila (Celotto and Graveley 2001).
Unlike promoter activity that predominantly regulates the abundance of transcripts, alternative splicing influences the structure of the mRNAs and their encoded proteins. As a result, it influences binding properties, intracellu-lar localization, enzymatic activity, protein stability, and post-translational modification of numerous gene products (Stamm et al. 2005). The magnitude of the changes evoked by alternative splicing are diverse and range from a complete loss of function to very subtle, hard to detect effects (Stamm et al., 2005). Alternative splicing can indirectly regulate transcript abundance.
University of Erlangen, Institute for Biochemistry, Fahrstraße 17, 91054 Erlangen, Germany
Progress in Molecular and Subcellular Biology Philippe Jeanteur (Ed.) Alternative Splicing and Disease © Springer-Verlag Berlin Heidelberg 2006
About 25-35% of alternative exons introduce frameshifts or stop codons into the pre-mRNA (Stamm et al. 2000; Lewis et al. 2003). Since approximately 75% of these exons are predicted to be subject to nonsense-mediated decay, an estimated 18-25% of transcripts will be switched off by stop codons caused by alternative splicing and nonsense mediated decay (Lewis et al. 2003). Finally, several proteins that regulate splice-site usage shuttle between nucleus and cytosol where they regulate translation (Sanford et al. 2004).
Splice Sites are Selected Through Combinatorial Control
Proper splice site selection is achieved by binding of protein and protein: RNA complexes (trans-factors) to weakly defined sequence elements (cis-factors) on the pre-mRNA (Fig. 1A). Binding of the trans-factors occurs cotranscriptionally and prevents the pre-mRNA from forming RNA:DNA hybrids with the genomic DNA. RNP complexes forming around exons promote binding of U2AF and U1 snRNP at the 3' and 5' splice sites respectively, which marks the sequences to be included in the mRNA. Sequences located in exons or the flanking introns can act as splicing silencers or enhancers. All cis-elements can only be described as consensus sequences that are loosely followed (Black 2003) and in general, they bind only weakly to trans-acting factors. The action of the ciselements depends on other surrounding elements, and due to this sequence context the same sequence can either promote or inhibit exon inclusion (Carstens et al. 1998). In order to achieve the high fidelity of splice site selection, multiple weak interactions are combined (Maniatis and Reed 2002; Maniatis and Tasic 2002) and as a result of this combinatorial control, splice site selection is influenced by multiple factors (Smith and Valcarcel 2000). This combinatorial control is mirrored in the complex composition of splicing regulatory complexes that often combine overlapping enhancing and silencing parts that collaborate to regulate exon usage (Singh et al. 2004b; Pagani et al. 2003b).
The formation of a specific protein:RNA complex from several intrinsically weak interactions has several advantages: (1) it allows a high sequence flexibility of exonic regulatory sequences that puts no constraints on coding requirements; (2) the protein interaction can be influenced by small changes in the concentration of regulatory proteins, which allows the alternative usage of exons depending on a tissue and/or developmental-specific concentration of regulatory factors; (3) phosphoryla-tion of regulatory factors that alter protein:protein-interactions can influence splice site selection; (4) the regulatory proteins can be exchanged with other proteins after the splicing reaction, allowing a dynamic processing of the RNA.
Fig. 1A, B. Change of splice site selection during disease. A: Formation of RNP complexes to recognize splice sites. The exon is shown as a gray square, the intron as lines. The formation of a complex between SR proteins and hnRNPs on two exonic enhancers (small boxes in the exons) is shown. This complex stabilizes the binding of U2AF to the 3' splice site and of U1snRNP to the 5' splice site of the exon (small lines show RNA:RNA binding). Multiple intrinsically weak protein:protein (red) interactions allow the formation of a specific complex. B: Mechanisms to change exon recognition. The formation of RNP complexes around exons can be disturbed by different ways. 1: Mutations in regulatory sequences can abolish binding of regulatory factors. 2: The concentration of regulatory factors can be altered, either by sequestration in different compartments or through a change of their expression level. 3: Phosphorylation events change the interaction between regulatory proteins, which interferes with exon recognition. Phosphorylation can either inhibit or promote protein:protein interaction. Only the inhibition is shown.
The usage of alternative exons changes during development or cell differentiation both in vivo and in cell cultures. Furthermore, numerous external stimuli have been identified that change alternative splicing patterns. In most cases, these changes are reversible, indicating that they are part of a normal physiological response (Stamm 2002).
Was this article helpful?