Science Wiki

Alternative Splicing (or differential splicing) is the regulated co-transcriptional and post-transcriptional process in eukaryotes by which the exons and introns of a primary precursor messenger RNA (pre-mRNA) transcript are joined in multiple combinatorial arrangements by the spliceosome, allowing a single genomic DNA locus to encode two, dozens, or even thousands of structurally and functionally distinct mature mRNA variants and protein isoforms. When the International Human Genome Sequencing Consortium completed the reference human genome in 2001–2004, biologists were astonished to discover that *Homo sapiens* possesses only ~19,900 to 20,500 protein-coding genes—roughly the same gene count as the microscopic roundworm *Caenorhabditis elegans* (~20,000 genes) and fewer than rice (*Oryza sativa*). High-throughput transcriptome sequencing (RNA-seq) resolved this C-value / G-value paradox by proving that approximately 95% of all multi-exon human genes undergo alternative RNA splicing, expanding the human proteome to well over 100,000 distinct functional polypeptide isoforms tailored to specific tissues, developmental stages, and physiological signals.

Whereas constitutive splicing joins every exon of a pre-mRNA in strict linear order (Exon 1 – Exon 2 – Exon 3) without exception, alternative splicing occurs whenever a splice site is suboptimal or regulated, forcing two or more splice junctions to compete for spliceosome assembly. All complex alternative splicing patterns are constructed from five fundamental binary modes (plus alternative promoter and polyadenylation choices):

Mechanistic Enforcement of Mutually Exclusive Exons

How does the cell guarantee that mutually exclusive exons (such as Exon 9 vs. Exon 10 in *PKM*, or the tandem exon clusters in *Drosophila Dscam1*) are never accidentally spliced together into the same mRNA? Cells use four biophysical and surveillance safeguards:

1. Steric Branch-Point Interference: In many mutually exclusive exon pairs (such as α-tropomyosin Exons 2 and 3), the branch point adenosine of the intervening intron is located anomalously close (<40–50 nucleotides) to the upstream 5' splice site. If U1 snRNP binds the 5' splice site of Exon 2, it sterically blocks U2 snRNP from binding the branch point of Exon 3, physically preventing the two exons from being ligated together!
2. Incompatible Spliceosome Types (U2 vs. U12): One internal splice junction may utilize a major U2-type GU donor while the adjacent junction requires a minor U12-type AC acceptor; because a hybrid U2–U12 intron cannot be spliced, the two exons cannot join.
3. Intronic RNA Secondary Structure ("Docking Site–Selector Site" Base Pairing): In the *Drosophila* Down syndrome cell adhesion molecule 1 (*Dscam1*) gene—which contains four mutually exclusive exon clusters (12 variants of Exon 4, 48 variants of Exon 6, 33 variants of Exon 9, and 2 variants of Exon 17)—a single conserved intronic docking site base-pairs competitively with only one complementary selector sequence upstream of a single variable exon in each cluster, enabling a single gene to generate 12 × 48 × 33 × 2 = 38,016 distinct cell-surface receptor isoforms that give every developing neuron a unique self-avoidance barcode!
4. Frame-Shift NMD Coupled Filtering: If both exons have lengths that are not multiples of 3 nucleotides (e.g., each is 100 nt), including a single exon preserves the reading frame (or combines with flanking frame), whereas including both or neither shifts the downstream reading frame and triggers destruction via Nonsense-Mediated mRNA Decay (NMD).

Because canonical 5' (GU) and 3' (AG) splice sites are short, degenerate motifs that appear by pure chance every few hundred nucleotides across long eukaryotic introns ("pseudo-exons"), the spliceosome relies on auxiliary *cis*-acting sequence elements—Splicing Enhancers and Splicing Silencers—bound by antagonistic families of *trans*-acting RNA-binding proteins:

Cis-Regulatory Elements and Trans-Acting Factors

• Exonic Splicing Enhancers (ESEs) and Intronic Splicing Enhancers (ISEs): Short (6–8 nt) purine-rich or AC-rich RNA motifs located inside an exon or flanking intron that promote exon inclusion. ESEs are bound primarily by SR proteins (Serine/Arginine-rich splicing factors, such as SRSF1 [ASF/SF2], SRSF2 [SC35], and SRSF3). SR proteins possess one or two N-terminal RNA Recognition Motifs (RRMs) that bind the ESE and a C-terminal RS domain (rich in alternating phosphorylated arginine–serine dipeptides). The phosphorylated RS domain acts as a molecular bridge that directly recruits U1 snRNP to the downstream 5' splice site and U2AF35/U2AF65 to the upstream 3' polypyrimidine tract—a process termed exon definition.
• Exonic Splicing Silencers (ESSs) and Intronic Splicing Silencers (ISSs): Sequence motifs that repress splice site usage and cause exon skipping. ESS and ISS elements are bound primarily by heterogeneous nuclear ribonucleoproteins (hnRNPs)—such as hnRNP A1, hnRNP I (PTBP1, Polypyrimidine Tract-Binding Protein), and hnRNP L. hnRNPs repress splicing either by directly competing with U2AF65 for binding to the polypyrimidine tract, by multimerizing cooperatively along an exon to mask ESEs, or by dimerizing across two introns to loop out an entire cassette exon from the spliceosome.
• Tissue-Specific Master Splicing Regulators: Specialized cell types express dedicated RNA-binding proteins with position-dependent "RNA maps"—typically repressing an exon when bound upstream or inside it, and activating the exon when bound in the downstream intron. Examples include NOVA1/2 and RBFOX1/2/3 (NeuN) in the brain, RBM20 and MBNL1 (Muscleblind) in cardiac and skeletal muscle, and ESRP1/2 in epithelial cells.

Because over 80% of alternative splicing occurs co-transcriptionally while the pre-mRNA is still tethered to elongating RNA Polymerase II, alternative splicing is controlled not only by the static RNA sequence, but also by transcription elongation kinetics and chromatin structure through two complementary mechanisms:

1. The Recruitment Model: The phosphorylated C-terminal domain (CTD) of Pol II directly binds SR proteins and spliceosomal factors, while histone marks on nucleosomes (which are selectively enriched over exons compared to introns, and marked by H3K36me3 or H3K9me3) bind chromatin adapter proteins (such as MRG15 and HP1γ) that recruit splicing regulators like PTBP1 directly to nascent exons.
2. The Kinetic Competition ("Window of Opportunity") Model: Consider a cassette exon with a weak 3' splice site followed downstream by a constitutive exon with a strong 3' splice site. If RNA Polymerase II elongates slowly (paused by dense nucleosomes, intragenic DNA methylation, or a slow RPB1 mutant), the weak upstream exon emerges from the polymerase first and enjoys a long temporal "window of opportunity" to recruit U2 snRNP before the competing downstream strong splice site is even synthesized—favoring exon inclusion. Conversely, if Pol II elongates rapidly, both the weak and strong splice sites emerge almost simultaneously; the strong downstream site outcompetes the weak site, causing exon skipping!

Alternative splicing drives pivotal physiological switches across development, metabolism, and disease:

• The Warburg Effect in Cancer (*PKM1* vs. *PKM2*): The *PKM* gene encodes pyruvate kinase, the final rate-limiting enzyme of glycolysis, containing two mutually exclusive exons: Exon 9 and Exon 10. In differentiated adult muscle and brain, Exon 9 is included to produce constitutively active PKM1, driving oxidative phosphorylation in mitochondria. In proliferating embryonic cells and cancer cells, the oncogenic transcription factor c-MYC upregulates hnRNP A1, hnRNP A2, and PTBP1, which bind sequences flanking Exon 9 to repress it and force inclusion of Exon 10, generating PKM2—an allosterically regulated isoform that shunts glucose carbons into biosynthetic pathways for rapid tumor growth.
• The *Drosophila* Sex-Determination Cascade (*Sxl → tra → dsx*): In fruit flies, sex is determined entirely by a domino cascade of alternative splicing: an X:Autosome ratio of 1.0 in females activates the splicing regulator Sex-lethal (Sxl), which forces productive alternative 3' splicing of Transformer (*tra*), which in turn teams up with Tra2 (an SR protein) to direct female-specific alternative splicing of Doublesex (*dsx*^F) and repress male courtship splicing of Fruitless (*fru*^M).
• Antisense Oligonucleotide (ASO) Splicing Therapeutics: In Spinal Muscular Atrophy (SMA)—previously the leading genetic cause of infant mortality—patients carry homozygous deletions of *SMN1* and rely on the paralogous backup gene *SMN2*. However, a single silent C → T transition at position +6 of *SMN2* Exon 7 converts an ESE (bound by SRSF1) into an ESS (bound by hnRNP A1), causing ~90% of *SMN2* transcripts to skip Exon 7 and produce an unstable, degraded SMNΔ7 protein. Developed by Adrian Krainer and colleagues and approved in 2016 as nusinersen (Spinraza), an 18-nucleotide 2'-O-methoxyethyl (MOE) antisense oligonucleotide binds specifically to the Intronic Splicing Silencer ISS-N1 immediately downstream of *SMN2* Exon 7, displacing hnRNP A1/A2, restoring >90% Exon 7 inclusion, and rescuing motor neuron survival in infants.

Why do animals favor exon skipping while plants favor intron retention?

In vertebrates, genes evolved through "intron expansion," producing tiny ~150-nt exons separated by massive 5,000-nt introns. Because the spliceosome cannot reach across a 5,000-nt intron, vertebrates use exon definition (pairing U2AF at the 3' end of Intron 1 with U1 at the 5' end of Intron 2 *across the short exon*); therefore, failing to define an exon results in exon skipping. In plants and fungi, introns remain very short (~100–250 nt), so the spliceosome pairs sites directly across the intron (intron definition); failing to recognize a splice site leaves the short intron behind (intron retention).

How does alternative splicing coupled to Nonsense-Mediated Decay (AS-NMD) act as a genetic thermostat?

Roughly 15% to 30% of regulated alternative splicing events intentionally splice in a non-coding "poison exon" containing an in-frame Premature Termination Codon (PTC), targeting the mRNA for destruction by Nonsense-Mediated mRNA Decay (NMD). Notably, the genes encoding SR proteins and hnRNPs themselves contain ultra-conserved poison exons: when SRSF1 or PTBP1 protein levels rise too high in a cell, the excess protein binds its own pre-mRNA and splices in the poison exon (or skips an essential exon), destroying its own mRNA to maintain homeostatic feedback!

Why must symmetric cassette exons have a nucleotide length divisible by 3?

Because the genetic code is read by the ribosome in non-overlapping triplets (codons), inserting or skipping an internal cassette exon without shifting the downstream reading frame requires that both flanking introns lie in the exact same codon phase (Phase 0, Phase 1, or Phase 2)—meaning the exon's length in nucleotides must be an exact multiple of 3 (3*n*). Over 65% of conserved tissue-specific cassette exons in mammals are symmetric (3*n*), allowing modular domain swapping without frame-shifting.