Alternative splicing is a form of RNA splicing in which transcripts from the same gene are processed using different combinations of splice sites, producing distinct RNA isoforms. In protein-coding genes, these differences can generate messenger RNAs that encode different proteins, or change transcript stability and protein production without altering the encoded protein sequence. It is an important layer of gene expression regulation in eukaryotes, allowing a single genetic locus to produce multiple molecular outputs. However, the existence of an RNA isoform does not by itself establish that it produces a stable, functional protein. (nature.com)
Molecular basis
During transcription, genetic information is copied into a precursor RNA. Many protein-coding precursors contain exons, segments retained in a particular mature transcript, separated by introns, which are removed during splicing. Exons are not necessarily entirely protein-coding: they can also contain untranslated sequences. Alternative splicing changes the boundaries or combinations of retained segments rather than changing the underlying DNA sequence. (nature.com)
Most nuclear pre-mRNA splicing is performed by the spliceosome, a dynamic complex of proteins and small nuclear RNAs. Its components recognize the 5′ splice site, an internal branch-point sequence, and the 3′ splice site. Two successive chemical reactions form a branched, lariat-shaped intron intermediate and then join the flanking exons. Alternative splicing uses this same machinery; the difference lies principally in which sites are selected and paired. Structural studies have directly visualized spliceosome assembly and the catalytic intermediates underlying these reactions. (nature.com)
Major patterns
Several recurring patterns describe alternative splicing:
- Exon skipping: a cassette exon is included in one transcript but omitted from another.
- Mutually exclusive exons: one of two or more alternative exons is selected for a particular transcript.
- Alternative 5′ splice sites: different donor sites change the boundary at the upstream end of an intron.
- Alternative 3′ splice sites: different acceptor sites change the boundary at the downstream end.
- Intron retention: a sequence removed as an intron in one isoform remains in another.
These patterns can occur together, producing complex transcript structures. Genome-wide RNA-binding studies and sequencing analyses distinguish these event classes because their regulatory requirements and measurement properties differ. (nature.com)
Alternative first and last exons require particular care in classification. Their occurrence can involve splicing, but may also depend on different promoters or cleavage and polyadenylation sites. Alternative splicing is therefore one component of alternative RNA processing, not a synonym for every mechanism that generates transcript isoforms. Tissue-level studies have found coordinated variation in splicing and 3′-end processing. (nature.com)
Regulation
Splice-site selection depends on both RNA sequence and the cellular regulatory environment. Exonic and intronic sequences can function as splicing enhancers or silencers by recruiting RNA-binding proteins. These factors influence recognition of nearby splice sites and assembly of the spliceosome. Large-scale binding and perturbation experiments have shown broadly opposing effects of many serine/arginine-rich proteins and heterogeneous nuclear ribonucleoproteins, although individual effects depend on context. (nature.com)
The position of a regulatory protein’s binding site can determine whether it promotes or inhibits exon inclusion. For example, studies of Nova proteins in neurons showed that binding to particular exonic sequences could inhibit spliceosome recruitment, whereas binding to certain intronic sequences enhanced inclusion. Consequently, a factor cannot always be classified simply as an activator or repressor. (nature.com)
Splicing also interacts with transcription. Experiments using a slower RNA polymerase II demonstrated altered inclusion of specific alternative exons, supporting a kinetic connection between RNA synthesis and splice-site choice. Differences in regulatory-factor abundance and activity help establish tissue- and developmental-stage-specific splicing programs. (pubmed.ncbi.nlm.nih.gov)
Biological consequences
When alternative splicing changes coding sequences, it can alter amino acid composition, binding specificity, or other protein properties. A well-characterized example is the fruit-fly Dscam gene. Selection among four groups of alternative exons permits 38,016 possible isoforms, including 19,008 extracellular-domain combinations attached to either of two transmembrane segments. Experiments demonstrated preferential binding between matching isoforms. Individual cells express subsets of this potential repertoire, rather than necessarily producing every possible combination. (pmc.ncbi.nlm.nih.gov)
Other splice changes alter the reading frame or introduce a premature termination codon. Some resulting transcripts are degraded through nonsense-mediated mRNA decay, an RNA-surveillance pathway. Coupling alternative splicing to degradation can regulate protein abundance rather than expand protein diversity. Thus, “one gene, many proteins” captures only part of the process’s biological significance. (pubmed.ncbi.nlm.nih.gov)
Prevalence and investigation
Alternative splicing is widespread in humans. A 2008 deep-sequencing study of 15 human tissue and cell-line transcriptomes estimated that 92–94% of human genes underwent alternative splicing. Such estimates depend on sampling, sequencing depth, annotation, and detection criteria; they do not imply that every isoform is abundant or functional in every tissue. (nature.com)
Researchers investigate splicing using targeted reverse-transcription PCR, RNA-binding assays, and RNA sequencing. Short reads reveal exon–exon junctions, but can leave uncertainty about which distant events occur together in a complete transcript. Long-read sequencing can connect multiple splice junctions within individual RNA molecules. Comparative assessments show that isoform identification still depends on sequencing quality and computational methods. (nature.com)
A common event-level measure is percent spliced in (PSI), which estimates the proportion of relevant transcripts containing an exon or splice choice. Changes in PSI are distinct from changes in total gene expression. Establishing an isoform’s biological function additionally requires evidence beyond its detection, such as perturbation experiments or direct characterization of its protein product. (pmc.ncbi.nlm.nih.gov)