Transcription termination is the process that tells an RNA polymerase to stop copying DNA and release its newly made RNA strand. Without it, the molecular machinery that reads genes would barrel past the end of one gene and into the next, creating garbled transcripts, jamming up the genome, and wasting enormous amounts of cellular energy. Different organisms and different types of RNA polymerase use surprisingly different strategies to accomplish this seemingly simple task, and when termination fails, the consequences range from disrupted gene regulation to neurodegenerative disease.
How Bacteria Stop Transcription
Bacteria have two main ways to end transcription, and both have been studied in extraordinary detail because bacterial systems are simpler and more experimentally tractable than their eukaryotic counterparts.
The first route is called intrinsic termination (sometimes “Rho-independent” termination). It relies entirely on features encoded in the DNA sequence itself. As the RNA polymerase transcribes a terminator region, the freshly made RNA folds back on itself into a hairpin structure, a short stem-loop that forms just behind the enzyme. This hairpin physically disrupts the complex that holds the polymerase, the DNA, and the RNA together. The hairpin weakens contacts in the region where the RNA pairs with the DNA template, and it does so at a precise moment: when the polymerase has just transcribed a run of uracil residues (encoded by a stretch of thymine in the DNA). That U-rich stretch creates an unusually weak grip between the RNA and DNA, and it also causes the polymerase to pause, giving the hairpin extra time to fold. The combination of the hairpin pulling things apart and the weak U-rich hybrid holding things together poorly is enough to collapse the whole complex and release the RNA.1PubMed. The mechanism of intrinsic transcription termination Structural studies have captured snapshots of each stage of this process, showing how the polymerase pauses at the terminator, how the hairpin folds inside the enzyme itself, and how the enzyme then rewinds the transcription bubble to let go of both the RNA and the DNA.2PubMed Central. Structural basis for intrinsic transcription termination
The second route requires a protein called Rho. Rho is a ring-shaped enzyme that latches onto the RNA transcript and uses energy to chase down the polymerase. Researchers describe two modes of Rho action. In the “catch-up” mode, Rho threads the RNA through its ring and motors along it until it physically collides with the polymerase, forcing it off the DNA. In the “stand-by” mode, Rho is recruited near the polymerase by a factor called NusG and stays poised to engage the enzyme at specific sites or when particular signals arise.3PubMed Central. Rho-dependent transcription termination: mechanisms and roles in bacterial fitness and adaptation to environmental changes Both modes accomplish the same goal: prying the polymerase off the DNA and releasing the RNA.
Two accessory proteins, NusA and NusG, fine-tune both types of termination. NusA slows the polymerase down and reduces Rho-dependent termination efficiency, while NusG speeds the polymerase up and increases Rho-dependent termination.4PubMed. Combinatorial effects of NusA and NusG on transcription elongation and Rho-dependent termination in Escherichia coli The two factors also interact directly with each other, and NusA can actually pull NusG off the polymerase, giving the cell another layer of control over whether and when termination happens.5PubMed Central. Transcription is regulated by NusA:NusG interaction
How Eukaryotic Cells Terminate Protein-Coding Genes
In eukaryotes, RNA Polymerase II (Pol II) transcribes protein-coding genes, and the termination process is far more elaborate. Two competing models tried for years to explain how it works. The “allosteric” model proposed that something changes within the polymerase itself after it passes a signal in the RNA, slowing it down or destabilizing it. The “torpedo” model proposed that after the RNA is cut at its polyadenylation signal (the molecular address tag that marks where the mature messenger RNA should end), an enzyme called XRN2 latches onto the leftover dangling RNA still attached to the polymerase and chews through it like a molecular torpedo, catching up to the polymerase and knocking it off the DNA.
Current evidence points to both models being partially right. In human cells, the endonuclease CPSF73 first cleaves the RNA at the polyadenylation signal, and this cleavage event is absolutely required: eliminating CPSF73 causes massive readthrough transcription across the entire genome. After cleavage, CPSF73 creates the entry point for the XRN2 torpedo. But the polymerase also slows down in the termination region through the activity of a phosphatase called PP1, and inhibiting PP1 extends readthrough even when XRN2 is present. So the emerging picture is a combined mechanism: PP1 slows the polymerase down while XRN2 chases it from behind, and both contributions matter.6PubMed Central. A unified allosteric/torpedo mechanism for transcriptional termination on human protein-coding genes The cleavage and polyadenylation machinery that recognizes the signal in the RNA includes complexes called CPSF and CstF, which have been identified through genetic screens as essential players in this process.7PubMed Central. Genes involved in pre-mRNA 3′-end formation and transcription termination revealed by a lin-15 operon Muv suppressor screen
After cleavage at the polyadenylation site, Pol II does not stop immediately. Nascent RNA profiling data show that the polymerase continues transcribing on average about eight kilobases past the cleavage site before it finally dissociates. Genes typically have around four termination sites, and the median termination window spans roughly three kilobases. These termination zones tend to be GC-rich and overlap with positions where Pol II pauses or slows, reinforcing the idea that pausing helps the torpedo catch up.8PubMed Central. Nascent RNA Analyses: Tracking Transcription and Its Regulation
Promoter-Proximal Termination and the Integrator Complex
Not all termination happens at the end of a gene. In animal cells, a large fraction of Pol II molecules that begin transcribing never make it very far. After copying just a few dozen nucleotides, the polymerase pauses near the promoter, and at that point the cell faces a decision: release the polymerase into productive elongation, or shut it down. The Integrator complex is the machinery that shuts it down.
Integrator recognizes paused Pol II near the promoter and drives premature termination through two enzymatic activities: an endonuclease that cuts the short nascent RNA, and a phosphatase that strips away the chemical modifications the polymerase needs to enter productive elongation.9PubMed Central. Integrator is a global promoter-proximal termination complex This mechanism acts at essentially all Pol II-transcribed locations across the genome.10PubMed Central. Integrator endonuclease drives promoter-proximal termination at all RNA polymerase II-transcribed loci By preventing paused polymerases from moving forward, Integrator functions as a gene repression mechanism, conceptually similar to the transcription attenuation systems bacteria use. When the cell wants a gene expressed, the kinase P-TEFb releases the paused polymerase into productive elongation before Integrator can intervene. When it does not, Integrator terminates the attempt.11PubMed Central. The Integrator Complex Attenuates Promoter-Proximal Transcription at Protein-Coding Genes
Terminating Non-Coding RNA Transcripts
Not every RNA a cell makes encodes a protein. Small nuclear RNAs, small nucleolar RNAs, and a vast number of unstable transcripts from “pervasive transcription” all need to be terminated too, and they use a distinct pathway. In budding yeast, non-coding RNA transcription is terminated by the Nrd1-Nab3-Sen1 (NNS) complex. This complex recognizes specific sequence motifs in the nascent RNA, recruits the helicase Sen1 to unwind the RNA-DNA hybrid, and then hands off the released transcript to the exosome for processing or degradation.12PubMed Central. Molecular basis for coordinating transcription termination with noncoding RNA degradation Genome-wide mapping of where Nrd1, Nab3, and Sen1 bind to transcripts has confirmed their role in terminating and processing these non-coding RNAs throughout the yeast genome.13PLOS Genetics. Transcriptome-Wide Binding Sites for Components of the Saccharomyces cerevisiae Non-Poly(A) Termination Pathway: Nrd1, Nab3, and Sen1
This separation of labor is clean in yeast: protein-coding genes use the polyadenylation-dependent pathway, and non-coding RNAs use the NNS pathway.14PubMed Central. The Nrd1-Nab3-Sen1 transcription termination complex from a structural perspective The situation in higher eukaryotes is less tidy. Mammals lack direct equivalents of Nrd1 and Nab3, and the mechanisms they use to terminate non-coding transcripts are still being worked out, though the Integrator complex appears to handle some of this role.
How RNA Polymerase I and III Handle Termination
Eukaryotic cells have three nuclear RNA polymerases, and each has its own termination logic. Pol I transcribes the large ribosomal RNA precursor and relies on a DNA-binding protein called TTF-I (transcription termination factor for Pol I) that recognizes specific sequences called Sal boxes downstream of the ribosomal RNA genes. TTF-I acts as a roadblock: it binds to the DNA in the path of the oncoming polymerase and forces it to stop. The protein’s ability to bind DNA is normally masked by its own amino-terminal domain, which must be removed or remodeled for TTF-I to become active.15Nucleic Acids Research. The Amino-Terminal Domain of the Transcription Termination Factor TTF-I Causes Protein Oligomerization and Inhibition of DNA Binding At high concentrations, TTF-I can also form oligomers on its target sequence, with multiple molecules binding simultaneously to a single Sal box.16Nucleic Acids Research. Oligomerization of the Transcription Termination Factor TTF-I: Implications for the Structural Organization of Ribosomal Transcription Units
Pol III, which transcribes transfer RNAs and other small RNAs, has arguably the simplest termination mechanism of any eukaryotic polymerase. It can terminate on a short run of T residues in the non-template DNA strand without needing any protein factors at all.17PubMed Central. Mechanism of Transcription Termination by RNA Polymerase III Utilizes a Non-template Strand Sequence-Specific Signal Element However, this picture has turned out to be incomplete. Genome-wide studies combined with in vitro assays have shown that T-tracts are necessary but not always sufficient for termination, and that secondary structures in the nascent RNA can serve as important auxiliary signals.18PubMed Central. An integrated model for termination of RNA polymerase III transcription
Riboswitches and Antitermination
Bacteria have turned transcription termination into a versatile regulatory switch. One striking example involves riboswitches, segments of an mRNA that fold into specific shapes depending on whether a small molecule is bound to them. When the molecule is present, the RNA folds in a way that forms an intrinsic terminator hairpin, and transcription halts before the gene is fully copied. Without the molecule, the RNA folds differently, the hairpin cannot form, and transcription continues. For instance, when the metabolite S-adenosylmethionine binds to certain riboswitch RNAs, it directly triggers terminator formation and shuts off transcription of downstream sulfur-metabolism genes.19PubMed Central. The riboswitch-mediated control of sulfur metabolism in bacteria Other riboswitches, like those sensing magnesium or flavin, work through the Rho-dependent termination pathway: when the ligand is present, Rho gains access to the RNA and promotes termination.20PubMed Central. Riboswitch control of Rho-dependent transcription termination
Viruses have evolved the opposite trick. Bacteriophage lambda, when it infects a bacterial cell, needs to override the cell’s terminators to read its own late genes. It does this with antitermination proteins. The phage N protein binds to a small RNA hairpin near the start of the transcript and modifies the RNA polymerase so that it ignores terminators located far downstream, allowing full-length phage transcripts to be made.21PubMed Central. The antitermination activity of bacteriophage lambda N protein is controlled by the kinetics of an RNA-looping-facilitated interaction with the transcription complex A second phage protein, Q, takes a different approach: it catches the polymerase while it is pausing very early in a transcript, modifies the enzyme at that point, and then chases it past the pause site, producing a polymerase that transcribes straight through downstream terminators.22PubMed. Phage lambda gene Q antiterminator recognizes RNA polymerase near the promoter and accelerates it through a pause site
Why Termination Matters for Genome Organization
Termination does more than end individual transcripts. In compact genomes, where genes sit close together, efficient termination prevents one gene’s polymerase from crashing into the regulatory regions of its neighbor. In yeast, termination occurs soon after the polyadenylation signal precisely because neighboring genes are so tightly packed.23Molecular Cell. Eukaryotic transcription termination by RNA polymerase II DNA-binding proteins can also serve as physical roadblocks. In yeast, the protein Reb1p terminates read-through transcripts by sitting on the DNA and blocking the polymerase. Removing Reb1p leads to transcriptional interference between neighboring genes, demonstrating that these roadblock terminators are essential for keeping each gene’s expression under independent control.24PubMed. Roadblock termination by reb1p restricts cryptic and readthrough transcription
R-loops, structures in which the RNA hybridizes back to the DNA template while the non-template strand is displaced, also form at termination zones and appear to play a functional role in signaling the polymerase to stop. Evidence is mounting that cells use these “scheduled” R-loops as part of normal gene regulation, including transcription termination.25PubMed Central. Regulatory R-loops as facilitators of gene expression and genome stability
When Termination Fails in Disease
When the termination machinery breaks, the consequences can be severe. One well-studied example involves senataxin, a helicase encoded by the SETX gene. Senataxin is the human counterpart of yeast Sen1p and plays a direct role in resolving R-loops and supporting Pol II termination. Cells depleted of senataxin show defective transcription termination by Pol II, along with reduced binding of Pol II to certain genes and decreased transcription of those genes.26Human Molecular Genetics. Functional role for senataxin, defective in ataxia oculomotor apraxia type 2, in transcriptional regulation
Mutations in SETX cause distinct neurological diseases depending on the type of mutation. Loss-of-function mutations cluster around the helicase domain and cause ataxia with oculomotor apraxia type 2 (AOA2), a disorder characterized by progressive cerebellar degeneration and peripheral nerve damage. Gain-of-function mutations, in contrast, cause a form of amyotrophic lateral sclerosis called ALS4.27PubMed Central. Role of senataxin in R-loop-mediated neurodegeneration The AOA2 mutations disrupt senataxin’s interaction with the exosome, the RNA-degradation complex that normally cooperates with senataxin to resolve R-loops and repair the DNA damage they can cause. The ALS4 mutations do not affect this interaction.28Genes & Development. A SUMO-dependent interaction between Senataxin and the exosome, disrupted in the neurodegenerative disease AOA2, targets the exosome to sites of transcription-induced DNA damage What both disease classes share is a connection between faulty termination, unresolved R-loops, and neurodegeneration.
Viruses exploit termination failures too. Herpes simplex virus 1 (HSV-1) triggers widespread disruption of transcription termination at cellular genes while leaving its own viral genes unaffected. The result is that Pol II reads through the polyadenylation signals of host genes and continues transcribing for tens of thousands of nucleotides into downstream genes, creating aberrant read-through transcripts and even novel splicing between exons of neighboring genes. Hundreds of host genes appear to be transcriptionally active on the surface, but the read-through transcripts are not properly processed and are not translated into protein.29PubMed Central. Widespread disruption of host transcription termination in HSV-1 infection More broadly, various viral proteins interfere with the cleavage and polyadenylation machinery, causing read-through transcription and suppressing host gene expression as part of the “host shutoff” response that helps the virus commandeer the cell’s resources.30PubMed Central. Modulation of mRNA 3′-End Processing and Transcription Termination in Virus-Infected Cells
Transcription Termination in Archaea
Archaea are the third domain of life, and their transcription machinery is a fascinating hybrid: their RNA polymerase resembles the eukaryotic Pol II more than it resembles bacterial RNA polymerase, yet they often organize their genes into operons like bacteria do. Their termination systems reflect this in-between status. Archaeal genes frequently end with U-rich sequences, reminiscent of the bacterial intrinsic terminator, but the factor responsible for termination at these sites turns out to be a homolog of eukaryotic CPSF73, the same endonuclease that cleaves RNA at the polyadenylation signal in humans. This archaeal factor, called aCPSF1 or FttA, binds specifically to the terminator U-tract and uses that binding, together with its nuclease activity, to drive termination. The strength of the interaction scales with the length of the U-tract: longer U-tracts mean tighter binding and higher termination efficiency.31PubMed Central. aCPSF1 cooperates with terminator U-tract to dictate archaeal transcription termination efficacy
A second archaeal termination factor, called Eta, works differently. Eta uses energy to actively disrupt elongation complexes, releasing the RNA. It requires upstream DNA sequences and appears to specialize in clearing polymerases that have become stalled or arrested, somewhat analogous to Rho in bacteria.32PubMed Central. Factor-dependent archaeal transcription termination The fact that archaea have factors resembling both the bacterial Rho system and the eukaryotic CPSF73 system suggests that the basic strategies for terminating transcription may be deeply conserved across all cellular life, even as the specific molecular players have diverged considerably.