RNA processing: 5-prime capping, splicing, 3-prime polyadenylation, and alternative splicing

Intuition Beginner

When RNA polymerase finishes transcribing a gene, the raw RNA copy is not yet ready to be translated into protein. It is called pre-mRNA (pre-messenger RNA) and needs three processing steps before it can leave the nucleus.

First, a modified guanine nucleotide called the 5-prime cap is attached to the front of the RNA. Think of it like putting a return address on a letter: it marks the RNA as legitimate and helps the ribosome recognise it later.

Second, long stretches of non-coding sequence called introns are cut out, and the remaining coding segments called exons are joined together. This is splicing. A massive molecular machine called the spliceosome performs the cutting and joining with single-nucleotide precision.

Third, a tail made of many adenine nucleotides is added to the end of the RNA. This poly-A tail protects the mRNA from degradation and helps regulate how long it lasts in the cell.

A single gene can produce multiple different proteins through alternative splicing. By joining different combinations of exons together, the cell generates distinct mRNA molecules from the same gene. This is one reason the human proteome (all proteins) is far larger than the human genome (all genes).

Visual Beginner

The diagram shows the three processing steps applied to a freshly transcribed pre-mRNA. The 5-prime cap (a modified G nucleotide) is added first, while transcription is still underway. Splicing removes introns and joins exons. The poly-A tail is added after cleavage at the 3-prime end. The finished mRNA then exits the nucleus for translation.

Worked example Beginner

A pre-mRNA has the structure: Exon 1 - Intron A - Exon 2 - Intron B - Exon 3.

After splicing, both introns are removed and the three exons are joined:

Step	Molecule
Pre-mRNA	Exon1--IntronA--Exon2--IntronB--Exon3
After splicing	Exon1--Exon2--Exon3

If alternative splicing skips Exon 2, the mature mRNA becomes Exon1--Exon3, encoding a shorter protein with a different function. A real example: the human troponin T gene produces different isoforms in heart vs. skeletal muscle by including or skipping specific exons, producing proteins with distinct contractile properties suited to each tissue.

Check your understanding Beginner

Formal definition Intermediate+

Eukaryotic pre-mRNA undergoes three co-transcriptional processing events that convert the primary transcript into translatable mRNA. All three begin while RNA polymerase II is still actively transcribing the gene.

5-prime capping

The 5-prime cap is a 7-methylguanosine ($m^{7}G$) linked to the first transcribed nucleotide by an unusual 5-prime-to-5-prime triphosphate bridge. Capping occurs in three enzymatic steps:

1. RNA 5-prime triphosphatase removes the gamma-phosphate from the 5-prime triphosphate of the nascent transcript.
2. Guanylyltransferase transfers GMP from GTP to the 5-prime diphosphate, forming the 5-prime-5-prime linkage.
3. Methyltransferase methylates the added guanine at the N-7 position, producing $m^{7}GpppN$.

In metazoans, the first ($2^{\prime }$-O-methyl) and sometimes second transcribed nucleotides are additionally methylated. Capping is coupled to the phosphorylated C-terminal domain (CTD) of RNA polymerase II and begins when the nascent transcript is approximately 20-30 nucleotides long.

Splicing and the spliceosome

Splicing removes introns and ligates flanking exons. Nearly all introns in nuclear pre-mRNA follow the GU-AG rule: introns begin with GU at the 5-prime splice site and end with AG at the 3-prime splice site. Three conserved sequences define an intron:

• 5-prime splice site consensus: AG|GURAGU (where | marks the exon-intron junction and R = purine)
• Branch point sequence: YNYURAY (where Y = pyrimidine and the underlined A is the branch point adenosine)
• 3-prime splice site consensus: YAG|R (polypyrimidine tract followed by YAG)

The spliceosome is a megadalton ribonucleoprotein (RNP) complex composed of five small nuclear ribonucleoproteins (snRNPs) — U1, U2, U4, U5, and U6 — plus numerous non-snRNP protein factors. Assembly proceeds stepwise through defined intermediates:

1. Complex E (commitment complex): U1 snRNP base-pairs with the 5-prime splice site; SF1 (splicing factor 1) binds the branch point; U2AF (U2 auxiliary factor) binds the polypyrimidine tract and 3-prime splice site.
2. Complex A (pre-spliceosome): U2 snRNP replaces SF1 at the branch point, bulging out the branch point adenosine. ATP-dependent RNA helicases drive this rearrangement.
3. Complex B (pre-catalytic spliceosome): The U4/U6.U5 tri-snRNP joins. U6 is base-paired with U4 and is inactive. U5 bridges the two exons.
4. Complex $B^{\ast }$ (activated spliceosome): U1 and U4 are displaced. U6 replaces U1 at the 5-prime splice site and pairs with U2, forming the catalytic core. This rearrangement requires the Brr2 helicase to unwind U4/U6.
5. Complex C (catalytic spliceosome): The two transesterification reactions occur (see Key mechanism below).

3-prime cleavage and polyadenylation

The 3-prime end of pre-mRNA is generated by endonucleolytic cleavage followed by polyadenylation. The key signal is the polyadenylation signal AAUAAA (or close variants), located 10-30 nucleotides upstream of the cleavage site.

The processing complex includes:

• CPSF (cleavage and polyadenylation specificity factor): recognises AAUAAA via CPSF-160 and the recently discovered CPSF-30/ZFC3H1 complex.
• CstF (cleavage stimulation factor): binds the GU-rich downstream sequence element (DSE).
• CF Im and CF IIm (cleavage factors): recognise upstream UGUA motifs and participate in cleavage.
• PolyA polymerase (PAP): adds the poly-A tail after cleavage.

After cleavage, PAP adds approximately 200-250 adenine residues (in mammals) to the new 3-prime end. The initial addition is slow and distributive; after approximately 10-12 A residues, PABPN1 (nuclear poly-A binding protein) binds the short tail and stimulates rapid, processive addition. PABPN1 also determines tail length by measuring the distance from the poly-A signal.

Alternative splicing

Alternative splicing generates multiple mRNA isoforms from a single pre-mRNA. The major modes are:

• Exon skipping (cassette exons): An exon is either included or excluded. This is the most common mode in mammals (~40% of alternative splicing events).
• Alternative 5-prime splice sites: Two competing 5-prime splice sites in the same intron produce exons with different 3-prime ends.
• Alternative 3-prime splice sites: Two competing 3-prime splice sites produce exons with different 5-prime ends.
• Mutually exclusive exons: Exactly one of two consecutive exons is included; both are never included together.
• Intron retention: An intron remains in the mature mRNA rather than being spliced out. More common in plants and lower eukaryotes than in mammals.

Regulation is mediated by cis-acting elements (exonic and intronic splicing enhancers or silencers, ESEs/ISEs/ESSs/ISSs) and trans-acting factors (SR proteins as enhancers, hnRNP proteins as silencers). The balance of these factors, which varies by cell type and condition, determines splice-site selection.

Key mechanism Intermediate+

The two transesterification reactions of splicing.

Splicing proceeds through two sequential transesterification reactions, both catalysed by the spliceosome's RNA components (primarily U2 and U6 snRNA). No high-energy cofactor (ATP, GTP) is consumed in the chemistry of bond breaking and bond formation — the reactions are isoenergetic rearrangements. ATP is required only for the helicase-driven conformational rearrangements between steps.

Step 1: Branching (lariat formation).

The 2-prime hydroxyl of the branch point adenosine performs a nucleophilic attack on the phosphodiester bond at the 5-prime splice site. This breaks the exon 1--intron junction and forms a new 2-prime-5-prime phosphodiester bond between the branch point A and the first nucleotide of the intron (the G of GU). The products are: freed Exon 1 (with a 3-prime OH) and the intron-Exon 2 lariat intermediate (a circularised structure with a 2-prime-5-prime branch and a tail containing Exon 2).

Step 2: Exon ligation.

The 3-prime OH of the freed Exon 1 attacks the phosphodiester bond at the 3-prime splice site (the intron--Exon 2 junction). This joins Exon 1 and Exon 2 with a standard 3-prime-5-prime phosphodiester bond and releases the intron lariat. The lariat is subsequently debranched by a 2-prime-5-prime phosphodiesterase and degraded.

The catalytic magnesium ions are coordinated by conserved nucleotides in U6 snRNA (U6 positions U74, G78 in yeast; analogous positions in humans). Cryo-EM structures at 3-4 Angstrom resolution show that the spliceosome's catalytic core resembles group II intron self-splicing elements, supporting the evolutionary hypothesis that the spliceosome descends from an ancient group II intron ribozyme.

Exercises Intermediate+

Advanced treatment Master

The minor (U12-type) spliceosome

Approximately 0.5% of human introns are spliced by the minor spliceosome, which uses U11, U12, U4atac, U5, and U6atac snRNPs instead of U1, U2, U4, U5, and U6. Minor introns have distinct consensus sequences (AU-AC at the splice sites instead of GU-AG, though some are GU-AG with different flanking sequences) and are enriched in genes involved in DNA repair, RNA processing, and voltage-gated ion channels. The minor spliceosome shares U5 snRNP with the major spliceosome but uses paralogous snRNAs for the other components. Mutations in U4atac snRNA or in the minor spliceosome protein RNPC3 cause microcephalic osteodysplastic primordial dwarfism type I (MOPD1), highlighting the nontrivial role of minor intron splicing in human development.

Cryo-EM structures of the spliceosome

A revolution in spliceosome structural biology occurred between 2015 and 2019, when cryo-electron microscopy resolved the spliceosome at near-atomic resolution across multiple states: the pre-B, B, $B^{act}$, $B^{\ast }$, C, C*, P, and intron-lariat spliceosome (ILS) complexes. The Shi, Zhang, and Luhrmann laboratories independently determined structures of human and yeast spliceosomes at 3.0-4.5 Angstrom resolution.

Key findings from cryo-EM:

• The catalytic core is formed by U2 and U6 snRNAs, with two coordinated magnesium ions ($M{g}^{2+}$) positioned to stabilise the leaving groups and incoming nucleophiles during both transesterification reactions. The active-site geometry closely matches that of group II intron ribozymes.
• The spliceosome undergoes massive conformational rearrangements between states. Between Complex B and Complex $B^{act}$, the U1 and U4 snRNPs dissociate and the U6 intramolecular stem-loop (ISL) folds into the catalytic conformation. Between Complex C and Complex C*, the spliceosome rotates the 3-prime splice site into the active site while the lariat remains anchored.
• The PRP8 protein, the largest and most conserved spliceosomal protein (~280 kDa), forms a scaffold that cradles the catalytic RNA core. Mutations in PRP8 cause retinitis pigmentosa type 13 (RP13), likely by destabilising the catalytic centre.
• Step 1 and Step 2 occur in distinct conformations of the same spliceosome, with the active site being reconfigured between the two reactions by the helicases Prp2 (before Step 1) and Prp16 (between Steps 1 and 2). This "two-state" model explains how a single active site catalyses two different transesterification reactions.

RNA catalysis and the evolutionary origin of the spliceosome

The spliceosome's catalytic core is composed of RNA (U2 and U6 snRNAs), with proteins playing structural and regulatory roles. This parallels group II intron self-splicing, where the intron RNA folds into a conserved three-dimensional structure (six domains, with domain V containing the catalytic centre) and catalyses its own excision in vitro without any proteins. The structural similarity between the spliceosome active site and group V of group II introns supports the hypothesis that the spliceosome evolved from a group II intron that fragmented into separate snRNAs during eukaryogenesis.

An active debate concerns whether the spliceosome is truly an RNA enzyme (ribozyme) or a protein-RNA hybrid catalyst. The cryo-EM structures show magnesium ions coordinated primarily by RNA functional groups, and no protein side chains are positioned to participate directly in phosphotransfer chemistry. The weight of evidence favours RNA catalysis, with proteins (PRP8, Cwc25, Yju2) stabilising the active-site conformation rather than performing chemistry.

Exon definition vs. intron definition

In vertebrates, where exons are short (~150 nt) and introns are long (up to hundreds of kilobases), splice-site recognition proceeds by exon definition: U1 snRNP binds the downstream 5-prime splice site and U2AF binds the upstream 3-prime splice site, with the interaction between these complexes occurring across the exon. After exon definition, a conformational rearrangement reconfigures the interactions to span introns for catalysis.

In organisms with short introns and long exons (e.g., S. cerevisiae), splice-site recognition proceeds by intron definition: U1 and U2 snRNPs interact across the intron. The mode of recognition has consequences for splicing fidelity: exon definition makes vertebrate splicing sensitive to exon size (very short or very long exons are poorly recognised) and explains why most disease-causing splice mutations in humans affect exon recognition rather than catalysis.

Nonsense-mediated decay and the exon junction complex

During splicing, the exon junction complex (EJC) is deposited approximately 20-24 nucleotides upstream of each exon-exon junction on the mRNA. The EJC core consists of four proteins (eIF4AIII, MAGOH, Y14, MLN51) and serves as a binding platform for multiple factors that influence mRNA fate: mRNA export (via the TAP/p15 export receptor), translation (via the SKAR protein and the mTOR pathway), and mRNA surveillance.

The EJC plays a central role in nonsense-mediated decay (NMD), which degrades mRNAs containing premature termination codons (PTCs). During the pioneer round of translation, the ribosome displaces EJCs as it traverses the mRNA. If translation terminates at a normal stop codon (typically in the last exon, downstream of all EJCs), all EJCs have been removed. If termination occurs at a PTC upstream of one or more EJCs, the terminating ribosome communicates with the downstream EJC through the UPF proteins (UPF1, UPF2, UPF3), triggering SMG1-mediated phosphorylation of UPF1 and subsequent mRNA degradation. The "50-55 nucleotide rule" (PTCs located more than 50-55 nucleotides upstream of an exon-exon junction trigger NMD) reflects the positional relationship between the termination codon and the nearest downstream EJC.

SMN complex and spinal muscular atrophy

The biogenesis of snRNPs requires the SMN (survival of motor neuron) complex, which assembles the heptameric Sm protein ring ($SmB/B^{\prime },D1,D2,D3,E,F,G$) around the Sm site of each snRNA. The SMN complex recognises the Sm proteins after their initial binding to snRNA and catalyses the formation of the Sm core. Mutations in the SMN1 gene cause spinal muscular atrophy (SMA), an autosomal recessive neurodegenerative disease characterised by loss of alpha motor neurons in the spinal cord. SMA severity ranges from type I (most severe, infantile onset, death before age 2) to type IV (adult onset). The disease mechanism is a dose-dependent reduction in functional SMN protein: humans have a nearly identical paralogue, SMN2, that primarily produces a truncated, unstable protein due to a C-to-T transition in exon 7 that causes Exon 7 skipping. A small fraction of SMN2 transcripts (~10-15%) include Exon 7 and produce functional protein.

Splice-switching therapeutics

Nusinersen (Spinraza), approved by the FDA in 2016 for SMA, is an antisense oligonucleotide (ASO) that binds the intronic splicing silencer N1 (ISS-N1) in intron 7 of the SMN2 pre-mRNA. By blocking the binding of splicing repressor proteins (particularly hnRNP A1/A2), nusinersen promotes inclusion of Exon 7 in SMN2 mRNA, increasing production of full-length functional SMN protein from the SMN2 gene. This was the first FDA-approved splice-switching therapy.

Other splice-switching ASOs include etelirsen (for Duchenne muscular dystrophy, promoting Exon 51 skipping in the dystrophin gene to restore the reading frame) and golodirsen (Exon 53 skipping). These therapies convert out-of-frame mutations into in-frame transcripts, producing shortened but partially functional dystrophin protein.

Beta-thalassemia splicing mutations

Beta-thalassemia provides a well-characterised example of how splice-site mutations cause disease. The most common beta-thalassemia mutations affect splice sites in the beta-globin gene (HBB):

• IVS1-110 G-to-A: A mutation at position 110 of Intron 1 creates a new AG dinucleotide that functions as a cryptic 3-prime splice site, causing aberrant splicing and a non-functional beta-globin. This is the most common beta-thalassemia mutation in Mediterranean populations.
• IVS1-1 G-to-A: A mutation at the first position of Intron 1 (the invariant G of the GT dinucleotide at the 5-prime splice site) abolishes normal splicing completely. This mutation results in beta-zero-thalassemia (no functional beta-globin produced), requiring lifelong transfusion therapy.
• IVS1-6 T-to-C: A mutation at position 6 of Intron 1 creates a cryptic 5-prime splice site. Some correctly spliced mRNA is still produced, resulting in beta-plus-thalassemia (reduced but not absent beta-globin).

These mutations demonstrate that the precise consensus sequences at splice sites are not merely informational — their integrity is essential for normal haemoglobin production and human health.

Recursive splicing

Some long introns in Drosophila and mammals are removed by recursive splicing, a process in which the intron is split into smaller segments that are sequentially excised. Recursive splice sites (RS sites) contain overlapping 3-prime and 5-prime splice site motifs: the AG at the end of one splicing event also serves as the GU context for the next (after a "zero-length exon" or "RS exon" of just a few nucleotides). Recursive splicing was discovered in the Drosophila Ultrabithorax (Ubx) gene and has since been identified in hundreds of human genes with long introns, including genes involved in neurodevelopment. The mechanism allows the spliceosome to process introns that are too long for a single round of definition and recognition.

Connections Master

• Transcription 17.05.02. RNA processing is co-transcriptional: the phosphorylated CTD of RNA polymerase II serves as a platform for recruiting capping enzymes, splicing factors, and 3-prime processing factors. The rate of transcription elongation affects splice-site selection — slower elongation gives weak splice sites more time to be recognised, a mechanism called "kinetic coupling." This means that transcription rate and splicing outcome are mechanistically linked.

• Translation 17.05.03. The 5-prime cap is the binding site for eIF4E, the poly-A tail recruits PABP, and together they circularise the mRNA to promote translation initiation. The EJCs deposited during splicing serve as markers for nonsense-mediated decay during the pioneer round of translation. Splicing, cap, tail, and translation are thus an integrated pipeline.

• Mutation and DNA repair 17.06.01. Splice-site mutations are among the most common causes of human genetic disease. Approximately 15-20% of disease-causing point mutations disrupt normal splicing, either by inactivating a splice site, creating a cryptic splice site, or altering an exonic splicing enhancer or silencer. Understanding splicing is essential for interpreting the clinical significance of genetic variants identified by whole-genome sequencing.

• Cell signalling 17.07.01. Alternative splicing is regulated by signalling pathways: the SR protein kinases (SRPK1, Clk/Sty) phosphorylate SR proteins in response to growth factor signalling, altering their activity and localisation. The mTOR pathway influences splice-site selection through the SKAR protein bound to EJCs. Signal-dependent splicing reprogramming occurs during differentiation, stress responses, and immune activation.

Historical notes Master

The discovery of split genes and RNA splicing in 1977 was entirely unexpected and upended the prevailing assumption that genes were contiguous stretches of DNA. Two groups independently made the discovery using electron microscopy and gel electrophoresis: Richard Roberts (working with adenovirus at Cold Spring Harbor Laboratory) and Phillip Sharp (working with adenovirus at MIT). When adenovirus mRNA was hybridised to the viral DNA template, the electron micrographs showed unexpected looped-out structures — the DNA contained sequences that were absent from the mature mRNA. These loops were the introns. Roberts and Sharp shared the 1993 Nobel Prize in physiology or medicine.

The term "intron" (intragenic region) and "exon" (expressed region) were coined by Walter Gilbert in 1978. Gilbert recognised that the existence of introns had profound implications for evolution: exons could be shuffled between genes to create new proteins, and alternative splicing could generate diversity from a single gene.

The spliceosome proved elusive biochemically because of its large size, dynamic composition, and low abundance. Christine Guthrie and John Abelson pioneered the genetic dissection of the spliceosome in Saccharomyces cerevisiae in the 1980s, using temperature-sensitive mutants that blocked splicing at specific steps. Their work identified the snRNAs and many of the protein factors. Reinhard Luhrmann and Claus Will performed the biochemical complement, purifying spliceosomal complexes from HeLa cells and systematically cataloguing their protein components — work that eventually identified over 300 distinct spliceosome-associated proteins.

The cryo-EM revolution in spliceosome structural biology (2015-2019) was enabled by advances in direct electron detectors and image processing algorithms. Yigong Shi (Tsinghua University), Rui Zhao (University of Colorado), and Clemens Plaschka, Maximilian Spenkuch, and Reinhard Luhrmann (Max Planck Institute) independently determined near-atomic structures of the spliceosome in multiple catalytic states, providing the first direct view of the RNA-based active site and the conformational rearrangements between steps.

The discovery of alternative splicing's prevalence followed from the EST (expressed sequence tag) sequencing projects of the 1990s and the completion of the human genome in 2001, which revealed that the genome contained far fewer protein-coding genes than the number of known proteins. Gilbert's 1978 prediction that alternative splicing could resolve this "paradox" was borne out: high-throughput RNA sequencing (RNA-seq) studies now estimate that 95% of multi-exon human genes undergo alternative splicing.

Bibliography Master

Alberts, B. et al. — Molecular Biology of the Cell, 7th ed. (Garland Science, 2022). Ch. 6: How Cells Read the Genome: From DNA to Protein.

Lodish, H. et al. — Molecular Cell Biology, 9th ed. (W. H. Freeman, 2021). Ch. 8: RNA Processing.

Will, C. L. & Luhrmann, R. — Spliceosome structure and function. Nat. Rev. Mol. Cell Biol. 12 (2011) 453-468.

Sharp, P. A. — Split genes and RNA splicing. Cell 77 (1994) 805-815.

Gilbert, W. — Why genes in pieces? Nature 271 (1978) 501.

Fica, S. M. & Nagai, K. — Cryo-electron microscopy snapshots of the spliceosome: structural insights into a dynamic ribonucleoprotein machine. Nat. Rev. Mol. Cell Biol. 18 (2017) 633-644.

Shi, Y. — Mechanistic insights into precursor messenger RNA splicing by the spliceosome. Nat. Rev. Mol. Cell Biol. 18 (2017) 655-670.

Wahl, M. C., Will, C. L. & Luhrmann, R. — The spliceosome: design principles of a dynamic RNP machine. Cell 136 (2009) 701-718.

Leff, S. E., Rosenfeld, M. G. & Evans, R. M. — Complex transcriptional units: diversity in gene expression by alternative RNA processing. Annu. Rev. Biochem. 55 (1986) 1091-1117.

Pan, Q., Shai, O., Lee, L. J., Frey, B. J. & Blencowe, B. J. — Deep surveying of alternative splicing complexity in the human transcriptome by high-throughput sequencing. Nat. Genet. 40 (2008) 1413-1415.

Singh, N. N., Singh, R. N. & Androphy, E. J. — Modulating role of RNPC3 in spinal muscular atrophy. Adv. Exp. Med. Biol. 623 (2007) 37-47.

Dziembowski, A. et al. — Proteomic analysis of the human spliceosome reveals its complex composition and dynamics. EMBO J. 23 (2004) 3443-3454.