During mRNA processing, the cell removes introns, which are non-coding stretches of RNA that interrupt the sequences used to build proteins. The average human gene contains about 7.8 introns, and all of them must be cut out before the messenger RNA can be used. But intron removal is only one part of a larger editing process that also trims and modifies both ends of the RNA molecule.
Introns: The Major Sequences Removed
When a gene is first copied from DNA into RNA, the result is called pre-mRNA. This raw transcript contains two types of sequences woven together: exons, which carry the protein-coding instructions, and introns, which do not. The cell must cut out every intron and stitch the remaining exons together to produce a functional messenger RNA. In the human genome, 26,564 annotated genes contain over 207,000 introns total, so this is not a minor housekeeping step. It happens on a massive scale, for nearly every gene, every time that gene is activated.
A large molecular machine called the spliceosome carries out this work. It recognizes specific signal sequences at the boundaries where each intron meets the neighboring exons, then executes two chemical reactions in sequence. First, it cuts one end of the intron and loops it back on itself to form a lasso-shaped structure called a lariat. Then it cuts the other end and joins the two flanking exons together. The intron lariat is released as waste. In yeast, this entire process can finish within seconds of the intron being transcribed, often while the rest of the gene is still being copied. In other cases, splicing happens after the full transcript is complete.
What Happens to Removed Introns
Once an intron lariat is released, a dedicated enzyme called a debranching enzyme cuts open the lasso structure, converting it back into a linear piece of RNA. From there, the cell typically breaks it down with enzymes that chew through RNA, recycling the individual nucleotides for reuse in new molecules.
Not all intron material is pure waste, though. Some introns contain embedded sequences that get processed into small functional RNAs instead of being destroyed. For example, certain introns in the gene for ribosomal protein RPL17 harbor small nucleolar RNAs that help chemically modify ribosomal RNA. Other introns house precursors for microRNAs, tiny molecules that regulate gene activity. The microRNA cluster miR-302, found inside an intron of the LARP7 gene, plays a role in early embryonic development across vertebrates from zebrafish to humans. The production of these small RNAs depends on splicing: the intron has to be removed from the pre-mRNA before it can be further processed into its functional form.
Trimming the 3′ End
The back end of the pre-mRNA also undergoes a removal step. The cell recognizes a signal sequence (most often the six-letter code AAUAAA) near the end of the transcript, then an enzyme cuts the RNA at a specific site downstream. Everything past that cut point is discarded. After clipping, the cell adds a long tail of repeated adenine nucleotides, called a poly(A) tail, which protects the finished mRNA from degradation and helps it get exported from the nucleus. One notable exception: the mRNAs for histone proteins, which package DNA, skip the poly(A) tail entirely. Their 3′ ends are simply cleaved and trimmed by a couple of nucleotides.
5′ Cap Addition (Not Removal, but Part of Processing)
At the front end of the molecule, mRNA processing adds material rather than removing it. A modified guanine nucleotide is attached in a reversed orientation to form what is called a 5′ cap. This cap is built in three steps: first, a phosphate group is removed from the very tip of the raw transcript, converting it from a triphosphate to a diphosphate end. Then a guanine nucleotide is linked on. Finally, a methyl group is added to that guanine. The finished cap protects the mRNA from being chewed up by enzymes and serves as a recognition signal for the cell’s protein-building machinery. So while the 5′ end involves removing a phosphate group, the net effect is adding a protective structure rather than stripping sequence away.
Alternative Splicing: When Exons Get Removed Too
Introns are always removed, but the cell sometimes removes certain exons as well. This process, called alternative splicing, allows a single gene to produce multiple different proteins depending on which exons are kept or skipped. There are several patterns. In exon skipping, an entire exon is left out of the final mRNA along with its flanking introns. In mutually exclusive exon selection, the cell picks one exon from a set and discards the others. In alternative splice-site selection, the spliceosome cuts at a different position within an exon or intron, changing where the boundary falls. And in intron retention, an intron that would normally be removed is deliberately kept in the final transcript.
These variations mean that the “what gets removed” question has no single fixed answer for a given gene. The same pre-mRNA can be processed into different mature forms in different tissues or at different stages of development, producing proteins with distinct functions from one genetic blueprint.
Genes That Skip the Process Entirely
About 901 human genes have no introns at all. These intronless genes produce mRNAs that need no splicing. Many of them belong to large gene families like G protein-coupled receptors and olfactory receptors. Their mRNAs still receive the 5′ cap and undergo 3′ end processing, but they bypass the spliceosome entirely, which means their path from transcription to a finished mRNA is shorter and simpler.
A Rare Edit: Changing Individual Bases
Beyond cutting and trimming, mRNA processing can also modify individual letters in the sequence. In the most well-studied form, called A-to-I editing, an enzyme chemically converts an adenosine base to inosine by removing a specific amino group through a reaction called deamination. The cell’s machinery then reads inosine as if it were guanine, effectively changing the message encoded in the RNA without altering the underlying DNA. This type of editing is selective, targeting specific sites in specific transcripts, and can change which amino acid ends up in the finished protein or alter how the mRNA is regulated.