Polyadenylation primarily occurs in the cell nucleus, where a large protein complex clips the freshly made messenger RNA (mRNA) and adds a tail of adenine bases to its end. But the nucleus isn’t the only location. Polyadenylation also takes place in the cytoplasm under specific biological circumstances and inside mitochondria, where it serves a different purpose than it does in the nucleus. In bacteria, polyadenylation happens in the cytoplasm as well, though it typically marks RNA for destruction rather than protection.
Nuclear Polyadenylation: The Primary Site
In eukaryotic cells (everything from yeast to humans), the main site of polyadenylation is the nucleus. As RNA polymerase II transcribes a gene into a pre-mRNA strand, a massive protein assembly called the cleavage and polyadenylation complex (CPAC) recognizes a specific signal sequence on the growing transcript. That signal is usually the six-letter code AAUAAA, located roughly 10 to 35 nucleotides upstream of where the cut will happen. Some genes use a slightly different version of this signal, like UAUAAA, but the canonical AAUAAA is by far the most common.
Once the CPAC locks onto the signal, it performs a two-step reaction. First, it cuts the pre-mRNA, releasing it from the polymerase. Then the enzyme poly(A) polymerase (PAP) adds a string of adenine nucleotides to the newly exposed end, creating the poly(A) tail. In humans, the predominant form of this enzyme is called PAPα, which exists in at least six different splice variants, with PAP II being the most active. Specialized poly(A) binding proteins coat the growing tail as it’s built, helping control its length and guiding the finished mRNA toward export from the nucleus.
Polyadenylation and Transcription Termination
Nuclear polyadenylation doesn’t happen in isolation. It is tightly coupled to the termination of transcription. Cleavage and polyadenylation factors physically associate with RNA polymerase II as it moves along the DNA, riding on the tail-like C-terminal domain of the polymerase’s largest subunit. When the polymerase passes the polyadenylation signal, the CPAC cleaves the transcript. The leftover RNA still attached to the polymerase is then rapidly degraded, which helps the polymerase disengage from the DNA and stop transcribing. In other words, polyadenylation is the event that tells the cell “this gene is done being copied.”
Cytoplasmic Polyadenylation
Some mRNAs receive additional poly(A) tail extension after they’ve already left the nucleus. This process, called cytoplasmic polyadenylation, was first described in sea urchin embryos in the 1970s and has since been observed in the oocytes and early embryos of insects, amphibians, fish, and mammals. During early development, the cell’s nucleus is often transcriptionally silent, meaning no new mRNA is being made. The cell instead relies on stockpiled mRNAs that were made earlier and kept dormant with short poly(A) tails. When the cell needs to activate those mRNAs, it lengthens their tails in the cytoplasm, switching on translation.
The key player in this process is a protein called CPEB1, which binds to specific sequences in the mRNA’s regulatory region. When triggered by a phosphorylation signal, CPEB1 recruits a cytoplasmic poly(A) polymerase (a noncanonical form called GLD2 in humans) to extend the tail and activate the mRNA.
Cytoplasmic polyadenylation isn’t limited to embryos. In neurons, CPEB1 and polyadenylation factors are found in the cell body and in dendrites, the branching extensions that receive signals from other neurons. mRNAs encoding proteins involved in strengthening synaptic connections, such as calmodulin kinase II, are stored at dendrites with short poly(A) tails. When a synapse is stimulated during learning, the tails are extended locally, triggering on-the-spot protein production. This localized translation is thought to be essential for the remodeling of synapses that underlies memory formation.
Mitochondrial Polyadenylation
Mitochondria, the energy-producing compartments inside cells, have their own small genome and their own RNA processing machinery. Polyadenylation in human mitochondria is carried out by a dedicated enzyme called mtPAP (also known as PAPD1), which is the only poly(A) polymerase in humans that carries a mitochondrial targeting signal directing it into the organelle. It adds relatively short poly(A) tails of about 50 adenines, compared to the 200 or more typically added in the nucleus.
Mitochondrial polyadenylation serves functions that are quite different from nuclear polyadenylation. For the majority of human mitochondrial mRNAs, the poly(A) tail actually completes the stop codon. The gene itself only encodes the first one or two letters of the UAA stop signal, and polyadenylation fills in the missing A residues. The tail also influences mRNA stability and translation efficiency inside the mitochondrion. Interestingly, in plant mitochondria, polyadenylation has the opposite effect: it marks mRNAs for degradation, reflecting the bacterial ancestry of mitochondria.
Polyadenylation in Bacteria
Bacteria polyadenylate their RNA in the cytoplasm (they have no nucleus), but the role is largely reversed compared to eukaryotic nuclear polyadenylation. In bacteria like E. coli, poly(A) tails are added to the 3′ ends of mRNAs and their degradation fragments by poly(A) polymerase I (PAP I). These tails act as landing pads for exonucleases, enzymes that chew RNA from one end. Bacterial mRNAs often have stable stem-loop structures at their 3′ ends that block these exonucleases. The addition of a single-stranded poly(A) stretch gives the exonuclease something to grab onto, overcoming the roadblock and allowing the RNA to be broken down.
This degradation-promoting function was first demonstrated for a small regulatory RNA involved in plasmid replication. Since then, poly(A) tails have been detected on many types of bacterial RNA, including primary transcripts, processed RNAs, and partially degraded fragments. Some bacteria, like Streptomyces coelicolor, don’t even use a dedicated poly(A) polymerase. Instead, an enzyme called polynucleotide phosphorylase handles the tail-adding duties. There are exceptions to the “polyadenylation equals degradation” rule in bacteria: in a few cases, such as transcripts from the E. coli flagellar operon, polyadenylation actually stabilizes the RNA.
Alternative Polyadenylation Across the Genome
About 70% of human genes contain more than one polyadenylation signal, meaning the same gene can produce mRNAs with different 3′ ends depending on which signal the processing machinery uses. This phenomenon is called alternative polyadenylation, or APA. Most of these alternative sites sit within the 3′ untranslated region of the mRNA, the stretch after the protein-coding sequence. Choosing a closer site produces a shorter 3′ region, while a more distant site yields a longer one.
This matters because the 3′ untranslated region is packed with regulatory elements that affect how stable the mRNA is, how efficiently it gets translated into protein, and where in the cell it ends up. A shorter version might dodge a binding site for a regulatory molecule that would otherwise suppress the mRNA, effectively boosting protein production. Cells routinely shift their polyadenylation site choices in response to developmental cues, changes in cell type, or disease states, making APA a powerful layer of gene regulation that operates entirely at the level of RNA processing.