RNA Processing Cheat Sheet: 5′ Cap, Splicing & Poly(A) Tail

RNA processing turns an RNA polymerase II transcript (pre-mRNA) into mature mRNA inside the nucleus in three steps: a 7-methylguanosine (m7G) cap is joined to the 5′ end by a 5′-to-5′ triphosphate bridge once about 25-30 nucleotides have been made; the spliceosome (snRNPs U1, U2, U4, U5 and U6) cuts out introns at GU...AG boundaries and releases each one as a lariat; and the 3′ end is cleaved 10-30 nucleotides downstream of an AAUAAA signal, where poly(A) polymerase adds about 200 adenosines without a template. The cap and tail protect the mRNA and help start translation. The clinical hooks are splice-site mutations (β-thalassemia, spinal muscular atrophy, the vignette below) and anti-Smith antibodies in lupus, which target snRNP proteins.

The one-page cheat sheet

Memory hook: mRNA graduates with a CAP, gets SNIPped, then grows a TAIL.

1. The 5′ cap (m7G)

• Timing: the first modification. Alberts: the cap goes on "as soon as RNA polymerase II has produced about 25 nucleotides"; Brown's Genomes: before the transcript reaches 30 nucleotides.
• Three enzymes, all bound to the phosphorylated tail of RNA polymerase II: a phosphatase removes one phosphate from the 5′ end, a guanylyl transferase adds GMP in reverse (5′-to-5′) orientation, and a methyltransferase adds the methyl group that makes it 7-methylguanosine. The ribose of the first one or two transcribed nucleotides is also methylated.
• Jobs: in the nucleus the cap binds the cap-binding complex (CBC), which helps the RNA get processed and exported. In the cytoplasm, eIF4E binds the cap and the small ribosomal subunit scans 5′→3′ to the first AUG.
• Decay: most mRNAs first lose their poly(A) tail, then the cap, and then nucleases digest them from both ends.
• Only RNA polymerase II transcripts get this cap; rRNA (Pol I) and tRNA (Pol III) are made uncapped.

2. Splicing: GU-A-AG and the lariat

The sequences. In the vast majority of introns the first two nucleotides are 5′-GU (the donor, or 5′ splice site) and the last two are AG-3′ (the acceptor, or 3′ splice site). Upstream of the AG lie the branch-point A and a polypyrimidine tract. Mnemonic: GU-A-AG.

Assembly, in order:

1. U1 snRNA base-pairs with the 5′ splice site.
2. SF1 (branch-point binding protein) marks the branch site and U2AF binds the polypyrimidine tract and 3′ splice site; U2 snRNA then base-pairs with the branch point.
3. The U4/U6 and U5 snRNPs join.
4. U6 leaves U4 and displaces U1 at the 5′ splice site. U2, U5 and U6 snRNAs are the catalytic core, so the spliceosome works as a ribozyme, much like group II self-splicing introns.

Chemistry: two transesterifications.

1. The 2′-OH of the branch-point A attacks the 5′ splice site. The intron's 5′ end is joined to that A by a 2′-5′ phosphodiester bond, giving the lariat.
2. The upstream exon's free 3′ end attacks the 3′ splice site: the exons are joined and the intron leaves as a lariat.

Anti-Smith and anti-U1 RNP: snRNPs as autoantigens

• Anti-Sm antibodies bind the Sm (Smith antigen) proteins that form the core of spliceosomal snRNP particles. They are highly specific for SLE but not sensitive, like anti-dsDNA; the sensitive screen is ANA, positive in more than 95% of patients with SLE.
• Anti-RNP (U1 RNP) occurs in SLE and in mixed connective tissue disease (MCTD), the classic exam pairing. The full antibody panel is in the SLE guide.

3. Cleavage and polyadenylation: the AAUAAA signal

• Signal: AAUAAA lies 10-30 nucleotides upstream of the polyadenylation site. Cleavage is often just after a CA, and a GU-rich element sits 10-20 nucleotides further downstream. CPSF (cleavage and polyadenylation specificity factor) binds the AAUAAA, and CstF (cleavage stimulation factor) binds the GU-rich element.
• Enzyme: poly(A) polymerase is a template-independent RNA polymerase. It adds about 200 adenosines (up to about 250); poly(A)-binding protein coats the tail.
• Timing: older sources call polyadenylation post-transcriptional, but it is part of how RNA polymerase II terminates transcription, which ends soon after the signal is transcribed.
• Jobs: the tail guards the 3′ end, since decay usually starts with tail shortening, and efficient translation initiation needs the tail bound by poly(A)-binding proteins.

One gene, several products

Alternative splicing, which is common in human genes, lets one gene make several proteins.

Splice-site mutations: what breaks

• A mutated donor GU or acceptor AG can't be recognized by U1 or U2AF. The intron may be retained, an exon skipped, or a nearby cryptic splice site used. The reading frame often shifts or meets a premature stop codon, giving unstable mRNA or a truncated protein. Compare all mutation types in point vs frameshift vs splice-site mutations.
• β-thalassemia: in GeneReviews (2026), β0 alleles (no β-globin) come from nonsense, frameshift and some splicing variants; β+ alleles (reduced β-globin) come from variants in introns, the promoter, the polyadenylation signal and the untranslated regions, as well as some splicing variants. See alpha vs beta thalassemia.
• Spinal muscular atrophy: SMA is caused by mutations in both copies of SMN1 on chromosome 5, which leave patients short of SMN protein. The similar SMN2 gene makes several SMN versions, but only one is full length and functional (the shorter ones are quickly degraded), and normally it supplies only 10-15% of functional SMN; more SMN2 copies usually mean milder, later-onset disease. Nusinersen, an intrathecal antisense oligonucleotide, increases exon 7 inclusion in SMN2 mRNA and so raises full-length SMN protein.

How USMLE tests RNA processing

For the transcription side (TFIID, promoters, enhancers) see transcription factors, and for what happens to the finished mRNA see translation initiation, elongation and termination.

Practice question

A 7-year-old boy has had recurrent pneumonia, sinusitis and otitis since infancy, plus chronic diarrhea. Serum IgG, IgA and IgE are low and IgM is normal. His activated T cells express almost no CD40 ligand. Sequencing of CD40LG shows a single-nucleotide change in the first two nucleotides of intron 3 (GT becomes AT in the DNA, so the intron starts with AU instead of GU in the pre-mRNA). Nuclear RNA from his T cells contains CD40LG transcripts that still include intron 3. Which step is most directly impaired?

A. Recognition of the 5′ splice site by U1 snRNP B. Addition of adenosines by poly(A) polymerase C. Addition of the 7-methylguanosine cap D. Binding of TFIID (TATA-binding protein) to the promoter E. Binding of the small ribosomal subunit to a Shine-Dalgarno sequence

Answer: A. The mutation destroys the GU donor at the start of intron 3. U1 snRNA finds the 5′ splice site by base pairing, so the intron is not removed and stays in the nuclear transcript. Little normal mRNA reaches the cytoplasm, CD40 ligand is missing from T cells, and B cells can't class switch: X-linked hyper-IgM syndrome (low IgG, IgA and IgE; normal or high IgM). More in the hyper-IgM flowchart.

• B: a poly(A) defect would leave the 3′ end unprotected and impair translation, but it would not leave one intron behind in otherwise processed transcripts.
• C: capping happens first, at about 25-30 nucleotides, and doesn't depend on intron sequences; a capping defect would also hit every RNA polymerase II transcript, not one gene.
• D: a promoter or TFIID problem lowers transcription, so there would be little CD40LG pre-mRNA at all, not abundant transcripts with an intron retained.
• E: Shine-Dalgarno sequences belong to bacterial mRNA; human ribosomes bind the 5′ cap via eIF4E and scan for AUG.

Key takeaways

• Cap first (at about 25-30 nucleotides); splicing and 3′ cleavage/poly(A) complete the job, all in the nucleus, for RNA polymerase II transcripts.
• Cap: m7G linked 5′-to-5′ through a triphosphate; protects the mRNA, helps export and recruits eIF4E for translation.
• Splicing: GU...A...AG; U1 binds the 5′ site, U2 the branch point; the 2′-OH of the branch A attacks first, making a lariat with a 2′-5′ bond.
• Poly(A): AAUAAA 10-30 nucleotides upstream of the cut; poly(A) polymerase needs no template and adds about 200 A's.
• Anti-Sm (snRNP protein) = highly specific, insensitive for SLE; anti-U1 RNP = MCTD.
• Alternative processing: calcitonin vs CGRP, membrane vs secreted antibody, apoB-48 by RNA editing.
• Splice-site mutations: intron retention, exon skipping or cryptic sites; seen in β-thalassemia, and splicing of SMN2 is the target of nusinersen in SMA.

Sources

• RNA Processing and Turnover — The Cell: A Molecular Approach, 2nd ed., NCBI Bookshelf (2000)
• From DNA to RNA — Molecular Biology of the Cell, 4th ed., NCBI Bookshelf (2002)
• From RNA to Protein — Molecular Biology of the Cell, 4th ed., NCBI Bookshelf (2002)
• Posttranscriptional Controls — Molecular Biology of the Cell, 4th ed., NCBI Bookshelf (2002)
• Synthesis and Processing of RNA — Genomes, 2nd ed., NCBI Bookshelf (2002)
• Biochemistry, Replication and Transcription — StatPearls, NCBI Bookshelf (2023)
• How to Design U1 snRNA Molecules for Splicing Rescue — Antisense RNA Design, Delivery, and Analysis, NCBI Bookshelf (2022)
• Systemic Lupus Erythematosus (SLE) — Merck Manual Professional Edition (2025)
• Calcium and Phosphate Homeostasis — Endotext, NCBI Bookshelf (2023)
• Beta-Thalassemia — GeneReviews, NCBI Bookshelf (revised 2026)
• SMN2 gene — MedlinePlus Genetics (2018)
• SPINRAZA (nusinersen) injection label — DailyMed (2026)