RNA Splicing: How Pre-mRNA Becomes Mature Messenger RNA

RNA splicing is the essential process that removes noncoding introns from precursor messenger RNA (pre-mRNA) and joins the remaining exons to produce a mature transcript ready for translation. This guide is for molecular biology researchers, clinical geneticists, bioinformaticians, and graduate students who need a clear, source-backed understanding of splicing mechanisms, splice site recognition, alternative splicing, and the careful interpretation of RNA sequencing (RNA-seq) data. By the end you will know how splicing works, why it matters for gene regulation and disease, and how to avoid common pitfalls when analyzing transcriptomic results. The NCBI Bookshelf provides a detailed introduction to the biochemistry of splicing NCBI Bookshelf.

Splicing is carried out by the spliceosome, a large ribonucleoprotein complex that recognizes specific sequences at exon intron boundaries. The accurate removal of introns and ligation of exons is required for the production of functional mRNA in eukaryotes. For anyone running or interpreting RNA-seq experiments, understanding splicing is not optional: nearly all multi exon genes undergo alternative splicing, and mis splicing underlies many genetic disorders. The EMBL EBI training materials offer practical tutorials on how splicing data are analyzed in bioinformatics workflows EMBL-EBI Training.

At a Glance

Introns, Exons, and Splice Sites

A gene transcribed by RNA polymerase II yields a pre-mRNA that contains both introns and exons. Introns can be much longer than exons and often harbor regulatory elements. The boundaries between an exon and an intron are defined by consensus sequences. At the 5' end of the intron is the donor splice site, almost always GU. At the 3' end is the acceptor splice site, almost always AG, preceded by a pyrimidine rich region. An invariant adenosine residue, the branch point, is located 18 to 40 nucleotides upstream of the acceptor site. These short motifs are necessary but not sufficient for correct splicing, additional exonic and intronic splicing enhancers and silencers help the spliceosome identify the correct sites. The NCBI Bookshelf chapter on RNA processing provides detailed diagrams of splice site sequences NCBI Bookshelf.

When a mutation alters a splice site, normal splicing can be disrupted. A recent study in Differentiation showed that a synonymous variant in the NPR2 gene creates a cryptic acceptor site, leading to aberrant pre-mRNA splicing and causing acromesomelic dysplasia A synonymous NPR2 variant causes acromesomelic dysplasia through aberrant pre-mRNA splicing. This illustrates that even a single nucleotide change without altering the amino acid sequence can have severe consequences.

The Splicing Machinery

The spliceosome assembles stepwise on the pre-mRNA. U1 snRNP binds the 5' splice site, U2 snRNP binds the branch point, and the U4/U6.U5 tri snRNP joins to form the active site. After rearrangements, U1 and U4 are released, and the catalytic center performs the first transesterification: the 2' hydroxyl of the branch point adenosine attacks the 5' splice site, forming a lariat structure. The second transesterification joins the 5' and 3' exons and releases the intron lariat, which is degraded. The excised lariat is debranched and the intron RNA is recycled or degraded. Alternative splicing is regulated by interacting RNA binding proteins that either promote or block the recruitment of spliceosomal components.

Splicing does not always occur between the same exon boundaries. A study in Life Science Alliance demonstrated that individual C. elegans neurons show differential sensitivity to smn-1 silencing, with distinct changes in alternative splicing patterns across neuronal subtypes Individual C. elegans neurons display differential sensitivity to smn-1 silencing. This highlights how splicing regulation can be cell type specific and how disruptions can produce tissue selective phenotypes.

Alternative Splicing

Alternative splicing generates multiple mature mRNAs from a single gene. The most common types are: exon skipping, mutually exclusive exons, alternative 5' splice sites, alternative 3' splice sites, and intron retention. Through these mechanisms, a single primary transcript can encode proteins with different domains, localization signals, or enzymatic activities. It is estimated that over 95% of human multi exon genes undergo alternative splicing, greatly expanding the proteomic diversity.

RNA binding proteins such as serine/arginine rich (SR) proteins and heterogeneous nuclear ribonucleoproteins (hnRNPs) bind to enhancer or silencer sequences and modulate spliceosome assembly. The combinatorial control of these factors allows spatial and temporal regulation of isoform expression. For example, the NRXN1 gene encodes neurexin proteins critical for synaptic function. Haploinsufficiency of NRXN1 was shown to converge on RNA processing during cortical development, with partial transcriptomic concordance seen in adult autism cortex NRXN1 haploinsufficiency converges on RNA processing during cortical development, with partial transcriptomic concordance in adult autism cortex. This study underscores how alternative splicing changes can be linked to neurodevelopmental disorders.

Why RNA-seq Interpretation Requires Care

Standard RNA-seq analysis maps short reads to a reference genome or transcriptome. If the aligner is not splice aware, reads that span exon exon junctions will not map correctly, leading to loss of information or false detection of deletions. Splice aware aligners such as STAR or HISAT2 use known splice junction annotations or detect junctions de novo. After alignment, quantification of gene and transcript abundances must account for reads that are compatible with multiple isoforms. Common tools like RSEM, Salmon, or kallisto use expectation maximization to resolve multi mapped reads.

Differential splicing analysis goes beyond simple gene level counts. Tools such as DEXSeq (available through Bioconductor) test for differences in exon usage, while rMATS or MAJIQ identify specific splicing events that change between conditions. The Galaxy Training Network provides hands on tutorials that walk users through the complete workflow from raw RNA-seq reads to differential splicing results Galaxy Training Network. The Bioconductor project hosts many R packages designed specifically for splicing analysis, including the widely used DEXSeq Bioconductor.

Because sequencing depth and read length affect the ability to detect splice junctions, researchers should carefully consider their experimental design. Short reads (50 bp) may not span junctions well, longer reads (100 150 bp) improve junction detection. Paired end sequencing greatly helps. The NCBI Sequence Read Archive (SRA) stores thousands of RNA-seq datasets that can be re analyzed to validate splicing results NCBI Sequence Read Archive.

Practical Workflow for Splice Aware Analysis

1. Quality control: Run FastQC on raw reads. Trim adapters and low quality bases with tools like Trimmomatic or Cutadapt.
2. Alignment: Use a splice aware aligner (STAR, HISAT2) with a reference genome and known annotation (GTF/GFF). Generate BAM files sorted by coordinate.
3. Junction detection: STAR outputs a SJ.out.tab file with detected splice junctions. Inspect the number of unique junctions and their read support.
4. Transcript quantification: Use Salmon or kallisto in alignment free mode or RSEM with BAM files to estimate transcript level abundances.
5. Differential splicing analysis: Use DEXSeq (exon level) or rMATS (event level). Provide a design matrix that accounts for batch effects.
6. Visualization: Plot exon usage changes with DEXSeq plots or sashimi plots using the SashimiPlot tool in IGV or via the ggsashimi R package.

A recent paper in Genome Research used a tissue and tumor atlas of RNA Polymerase III products to uncover context specific activities, showing the importance of precise transcriptomic mapping An RNA polymerase III tissue and tumor atlas uncovers context specific activities linked to 3D epigenome regulatory mechanisms. Their approach integrated splice aware quantification to separate Pol III transcripts from overlapping Pol II signals.

Common Mistakes

• Using a gene level count method for splicing analysis: If you sum all reads per gene, you lose isoform information. Always use transcript aware or exon level counts.
• Ignoring multi mapping reads: Many reads map to multiple isoforms. Simply discarding them biases results. Use expectation maximization methods.
• Assuming a single reference transcript per gene: Relying on the RefSeq canonical transcript may miss important isoforms. Use comprehensive annotations like GENCODE.
• Not validating with RT-PCR: Computational predictions of alternative splicing should be confirmed with experimental methods, especially for novel junctions.
• Overlooking batch effects: Splicing is sensitive to RNA quality and sequencing depth. Check for technical confounders in your differential analysis.
• Misinterpreting intron retention: It can be a genuine regulatory mechanism, but also a sign of incomplete splicing or genomic DNA contamination. Filter carefully.

Limits and Uncertainty

Our understanding of splicing regulation is incomplete. The rules that govern splice site choice are probabilistic, not deterministic. Many splicing events are tissue specific or developmentally regulated, and the effect of a mutation on splicing can be hard to predict without experimental validation. Computational prediction tools (e.g., SpliceAI, MaxEntScan) provide useful scores but have limited accuracy for cryptic or distant splice sites. The dynamic nature of the spliceosome means that changes in protein levels of splicing factors can have widespread effects. For example, the same synonymous variant in PCCA was shown to create a pseudoexon that could be targeted by CRISPR Cas12a gene editing, demonstrating both the challenge and the therapeutic opportunity Targeted gene editing of PCCA pseudoexon using CRISPR-Cas12a for potential therapy in propionic acidemia. Even with high quality RNA-seq, some low abundance isoforms may not be detected. Short read sequencing cannot always resolve complex isoform structures, long read technologies (PacBio, Oxford Nanopore) are increasingly used to obtain full length transcript sequences.

Frequently Asked Questions

What is a splice site and why is it important? A splice site is a short sequence at the boundary between an exon and an intron. The 5' splice site (donor) and 3' splice site (acceptor) are recognized by the spliceosome. Mutations that disrupt these sites often cause exon skipping or activation of cryptic sites, leading to disease.

How does alternative splicing lead to different proteins? By including or excluding specific exons, alternative splicing changes the mRNA coding sequence. This can add, remove, or replace protein domains, alter localization signals, or introduce premature stop codons. A single gene can produce dozens of isoforms.

Can RNA-seq detect all alternative splicing events? No. Short reads may not span long introns or complex rearrangements. Low coverage events may be missed. Detection also depends on annotation quality and the bioinformatics pipeline. Experimental validation remains important.

Do all eukaryotic genes undergo splicing? Most multi exon genes in vertebrates do, but some genes have only one exon and are not spliced. Splicing is nearly universal in eukaryotes, though intron density varies widely between species.

References and Further Reading

• NCBI Bookshelf: RNA Processing , Authoritative reference on splicing biochemistry and regulation.
• EMBL-EBI Training: RNA-seq analysis , Free courses on transcriptomics and differential splicing.
• Galaxy Training Network: Splice variant analysis , Hands on workflow for alternative splicing detection.
• Bioconductor: DEXSeq and other splicing packages , Open source software for differential exon usage.
• NCBI Sequence Read Archive (SRA) , Repository for raw sequencing data used in splicing studies.
• PubMed: NPR2 synonymous variant and aberrant splicing , Case study on how a silent mutation disrupts splicing.
• PubMed: C. elegans smn-1 silencing and neuronal splicing , Example of cell type specific alternative splicing.
• PubMed: NRXN1 haploinsufficiency and RNA processing in cortex , Linking splicing changes to neurodevelopmental disorders.
• PubMed: Pol III tissue and tumor atlas , Comprehensive transcript mapping with splice aware methods.
• PubMed: CRISPR editing of pseudoexon in PCCA , Therapeutic correction of an aberrant splicing event.

Related Articles

• RNA Vaccines: How mRNA Platforms Work and What They Require
• Types of RNA: Messenger, Transfer, Ribosomal, and Regulatory RNA Explained
• RNA Structure: Bases, Pairing, Folding, and Biological Function
• RNA Polymerase Function: How RNA Is Made From a DNA Template
• RNA Interference: A Practical Guide to Gene Silencing Mechanisms