RNA polymerase is the enzyme that reads a gene’s DNA sequence and builds a complementary strand of RNA from it, a process called transcription. It is the first physical step in turning genetic instructions into something a cell can use, whether that is a protein, a structural component of the ribosome, or a regulatory signal. Every living cell on Earth relies on some version of this enzyme, and all known multisubunit RNA polymerases descend from a single common ancestor, a detail that speaks to how fundamental the job is.1Nature Reviews Microbiology. Evolution of multisubunit RNA polymerases in the three domains of life But “reads DNA and makes RNA” is a dramatic undersell. The enzyme does not just copy information; it unwinds the double helix, proofreads its own work, navigates obstacles like tightly wound chromatin, and even helps recruit the cellular machinery that will process the RNA it produces.
Finding the Right Starting Point
Before a single RNA building block is joined to the next, RNA polymerase has to land on the correct stretch of DNA. In bacteria, the enzyme teams up with a helper protein called a sigma factor. Together, they scan the genome for specific DNA sequences known as promoters. Once the sigma factor recognizes a promoter, the polymerase clamps down and begins prying open the two DNA strands to create what researchers call an “open complex.” This is not a passive process. The bacterial RNA polymerase acts as a molecular machine that uses the energy of binding itself to physically remodel the DNA, separating the two strands around the start site so the template strand is positioned in the enzyme’s active site.2Journal of Molecular Biology. Mechanism of Bacterial Transcription Initiation: RNA Polymerase – Promoter Binding, Isomerization to Initiation-Competent Open Complexes, and Initiation of RNA Synthesis
The DNA sequence of the promoter itself has a surprising amount of influence over what happens next. Certain promoter sequences produce open complexes that are stable for hours, while others form complexes that last only seconds. Small sequence changes can reshape the transcription bubble globally, altering how fast the polymerase transitions from initiation into full-speed RNA synthesis.3PubMed Central. Structural origins of Escherichia coli RNA polymerase open promoter complex stability In eukaryotes, the picture is more elaborate: RNA Polymerase II, which transcribes protein-coding genes, cannot recognize a promoter on its own and requires a large team of general transcription factors to assemble at the gene’s start site before it can begin.
Escaping the Promoter
Getting started turns out to be the hardest part. After the open complex forms, RNA polymerase begins stitching together short RNA fragments, but it often fails to clear the promoter. The enzyme stays anchored to the promoter DNA while pulling downstream DNA into itself, a process called “scrunching.” This creates mechanical stress, like winding a spring. If the stress builds enough, it snaps the polymerase free of the promoter and the sigma factor, launching it into productive elongation. If not, the short RNA is released and the whole process starts over.4PubMed Central. Abortive initiation and productive initiation by RNA polymerase involve DNA scrunching
This “abortive initiation” is surprisingly common. In studies of E. coli RNA polymerase, fewer than half of open complexes successfully produce a full-length RNA transcript. Most stall after making a short RNA, release it, and try again.5PubMed Central. Mechanism of transcription initiation and promoter escape by E. coli RNA polymerase That sounds wasteful, but it acts as a checkpoint. Only complexes that generate enough scrunching energy push through, which ties the rate of gene expression to the physical properties of the promoter.
Building the RNA Chain
Once free of the promoter, RNA polymerase enters the elongation phase, and this is where the core catalytic work happens. The enzyme moves along the template strand of DNA, reading it one base at a time and adding the matching RNA nucleotide to the growing chain. Each addition follows a cycle: the incoming nucleotide enters the active site, a mobile element within the enzyme called the trigger loop folds into position to check that the nucleotide is correct, the chemical bond forms, and then the enzyme shifts forward by one position to accept the next nucleotide.6Nature Communications. Structural dynamics of RNA polymerase II throughout the nucleotide addition cycle The chemistry at the active site uses two metal ions to catalyze the bond between nucleotides, a mechanism shared across all multisubunit RNA polymerases.7PubMed Central. Nucleotide addition and cleavage by RNA polymerase II: Coordination of two catalytic reactions using a single active site
Recent work has zoomed in on this catalytic cycle at near-atomic resolution. Cryo-electron microscopy structures of yeast RNA Polymerase II at better than two-angstrom resolution revealed that hundreds of precisely positioned water molecules play active roles in the chemistry, helping shuttle protons during the bond-forming reaction and stabilizing the trigger loop as it folds.8PubMed Central. Sub-2 Å cryo-EM structures of transcribing RNA polymerase II reveal critical roles of water molecules in catalysis Water, in other words, is not just a passive solvent here; it is a participant in the reaction.
How RNA Polymerase Corrects Its Own Mistakes
RNA polymerase is not as accurate as the machinery that copies DNA, but it is not reckless, either. It has a built-in proofreading system that works very differently from DNA polymerase’s approach. When the wrong nucleotide gets incorporated, the mismatched base tends to fray away from the DNA template, which stalls the enzyme. The polymerase then slides backward along the DNA and RNA by one position, a movement called backtracking, which threads the mismatched end of the RNA into a secondary channel where it gets clipped off. The same active site used for adding nucleotides performs the cleavage, so RNA polymerase does not need a separate editing domain the way DNA polymerase does.9PubMed. RNA polymerase fidelity and transcriptional proofreading
Backtracking is not limited to error correction. RNA polymerase oscillates between forward and backward movements at many positions along a gene, making it more of a ratchet than a one-way train. These pauses and reversals play roles in elongation control, gene regulation, and even genome stability.10PubMed Central. RNA polymerase backtracking in gene regulation and genome instability If the polymerase backtracks too far, though, it can get stuck. Eukaryotic Pol I and Pol II handle this differently. Pol I has a built-in subunit that limits how far it slides and enables cleavage of RNA up to about 20 nucleotides deep. Pol II relies on an external helper, TFIIS, which can rescue it from backtracks of any depth by stimulating rapid RNA cleavage.11PubMed Central. Mechanisms of backtrack recovery by RNA polymerases I and II
Knowing When to Stop
Termination, the act of releasing the completed RNA and dissociating from the DNA, is just as tightly controlled as initiation. In bacteria, two main mechanisms handle this. Intrinsic termination relies on signals encoded directly in the DNA and the RNA transcript: a hairpin structure forms in the emerging RNA, followed by a string of uracils, and together these destabilize the elongation complex enough to cause the polymerase to fall off. Rho-dependent termination involves a separate protein, Rho, which latches onto the RNA and chases the polymerase, using energy from ATP to catch up and force the complex apart.12PubMed. Mechanisms of Bacterial Transcription Termination: All Good Things Must End A third, less common pathway uses a DNA-tracking motor protein called Mfd that can dislodge stalled polymerases by exerting torque on them.13PubMed. Mechanisms of Bacterial Transcription Termination
In eukaryotes, termination of protein-coding genes involves a partnership between two long-competing models that now appear to work together. After RNA Polymerase II transcribes past a poly(A) signal in the RNA, cleavage factors cut the transcript. The upstream piece becomes the functional messenger RNA, which gets a poly(A) tail added. The downstream piece is attacked by an exonuclease (called XRN2 in humans, Rat1 in yeast) that chews through the leftover RNA and catches up to the still-transcribing polymerase. Meanwhile, the polymerase itself slows down in the termination zone, partly through dephosphorylation of its tail. The exonuclease catching a slowed polymerase is what ultimately triggers release.14PubMed Central. A unified allosteric/torpedo mechanism for transcriptional termination on human protein-coding genes This combined model explains why both slowing the polymerase and degrading the downstream RNA are necessary for efficient termination.15Molecular Cell. Transcriptional termination by RNA polymerase II
Three Polymerases, Three Jobs
Bacteria get by with one RNA polymerase. Eukaryotes have at least three, each dedicated to a different class of genes. Pol I works in the nucleolus and transcribes the large ribosomal RNA precursor, the most abundant RNA in any cell. Pol II handles all protein-coding genes and many regulatory noncoding RNAs. Pol III makes transfer RNAs, the 5S ribosomal RNA, and a handful of other small RNAs.16PubMed. Structure of eukaryotic RNA polymerases All three share a structurally similar catalytic core, but they differ on their surfaces, which accounts for their distinct sets of transcription factors and promoter-recognition strategies.17PubMed Central. Functional divergence of eukaryotic RNA polymerases: unique properties of RNA polymerase I suit its cellular role
Plants go further. They have evolved two additional polymerases, Pol IV and Pol V, which diverged from Pol II. Neither one makes conventional transcripts. Instead, they produce noncoding RNAs used in gene silencing: Pol IV generates small RNA precursors that guide DNA methylation, and Pol V creates scaffold RNAs at silencing targets.18PubMed Central. RNA Pol IV and V in gene silencing: Rebel polymerases evolving away from Pol II’s rules Orthologs of both have been traced all the way back to early land plants, including mosses and liverworts, suggesting this silencing system is ancient and predates the evolution of flowering plants.19Molecular Biology and Evolution. Ancient Origin and Recent Innovations of RNA Polymerase IV and V
The CTD as a Coordination Platform
One of the most unusual features of RNA Polymerase II is its C-terminal domain, or CTD: a long, flexible tail on its largest subunit made of repeating seven-amino-acid units. Yeast has 26 of these repeats; vertebrates have 52. This tail does not participate directly in RNA synthesis. Instead, it acts as a landing pad for dozens of other proteins. As the polymerase moves through the transcription cycle, different amino acids in the CTD repeats get phosphorylated, creating a shifting pattern of chemical marks that recruit the right helpers at the right time: capping enzymes early on, splicing factors during elongation, and cleavage and polyadenylation machinery near the end.20PubMed Central. The RNA polymerase II CTD coordinates transcription and RNA processing The phosphorylation pattern is dynamic and changes across a gene’s length, so the CTD essentially broadcasts the polymerase’s position to the rest of the cell’s gene-expression machinery.21Biochimica et Biophysica Acta (BBA) – Gene Regulatory Mechanisms. Dynamic phosphorylation patterns of RNA polymerase II CTD during transcription
The CTD has another trick. Its repetitive, disordered structure lets it participate in liquid-like condensates inside the nucleus, clustering Pol II molecules together with transcription factors and coactivators at highly active genes and super-enhancers.22Nature Structural & Molecular Biology. RNA polymerase II clustering through carboxy-terminal domain phase separation Phosphorylation of the CTD during promoter escape appears to eject the polymerase from these clusters, setting it free for elongation. This means the same chemical modification that recruits processing factors also controls the polymerase’s physical location in the nucleus.
Pausing as a Regulatory Tool
On many eukaryotic genes, Pol II does not simply race through from start to finish. Shortly after beginning transcription, it pauses about 30 to 60 nucleotides downstream of the start site and waits. This promoter-proximal pausing is widespread and serves as a major control point for gene expression. The paused polymerase keeps the promoter region open and accessible, essentially placing the gene in a “ready” state. When the right signal arrives, a kinase complex releases the polymerase into productive elongation.23PubMed Central. Promoter-proximal pausing of RNA polymerase II: a nexus of gene regulation
This mechanism is especially important for genes that need to respond quickly to signals, such as those involved in immune responses, stress, and development. Because the polymerase is already loaded and has synthesized a short RNA, the gene can be fully activated in minutes rather than waiting for the entire initiation process to start from scratch.24PubMed Central. Ready, pause, go: regulation of RNA polymerase II pausing and release by cellular signaling pathways Recent cancer research has highlighted this checkpoint’s medical relevance: in acute myeloid leukemia cells, drugs that inhibit certain enzymes involved in chromatin modification can trigger release of paused Pol II, switching on genes that were poised but silent.25PubMed. HDAC Inhibition Triggers Release of RNA Polymerase II from Promoter-Proximal Pausing in Healthy Blood Progenitors and Primary Acute Myeloid Leukemia Myeloblasts
Transcribing Through Chromatin
In eukaryotes, DNA is wrapped around histone proteins into structures called nucleosomes, and these present a physical barrier to any enzyme trying to move along the DNA. Cryo-electron microscopy has captured RNA Polymerase II in the act of transcribing through a nucleosome, and the snapshots show the enzyme pausing at specific contact points where histones grip the DNA most tightly. At each of these sites, the polymerase gradually peels the DNA away from the histone surface while the histone core stays largely intact.26PubMed. Structural basis of the nucleosome transition during RNA polymerase II passage Elongation factors like SPT6 and the PAF complex help by sitting near the nucleosome and coordinating histone chaperones that can briefly hold displaced histones and replace them once the polymerase passes.27Nature Communications. Structure of transcribing RNA polymerase II-nucleosome complex
Without this escort system, transcription through chromatin would be slow and error-prone, and the chromatin landscape behind the polymerase would be disordered. The cell’s ability to maintain its epigenetic marks depends, in part, on RNA polymerase passing through without permanently disrupting nucleosome positioning.
RNA Polymerase as a DNA Damage Sensor
When RNA polymerase runs into a damaged base on the template strand, it stalls. That stalling event doubles as an alarm signal. In a process called transcription-coupled repair, the stalled polymerase recruits repair proteins to the damage site, prioritizing the fix of genes that are actively being transcribed. This makes biological sense: lesions in active genes are the most immediately threatening, and the polymerase’s inability to continue is the most reliable indicator of their location.28PubMed Central. Mechanistic insights into transcription coupled DNA repair
If repair does not happen quickly, the stalled polymerase itself becomes a problem. A stuck polymerase can block DNA replication, promote genome instability, or trigger cell death. Cells have a backup system for this: a pathway involving the protein p97 and the proteasome can physically remove the stalled polymerase from the DNA, even without repairing the lesion, as a last resort to prevent more widespread damage.29Nature Communications. Coordination of transcription-coupled repair and repair-independent release of lesion-stalled RNA polymerase II Both the repair-coupled and repair-independent pathways require the protein CSA and ubiquitination of the stalled polymerase, but the cell prioritizes actual repair when it can.30Nature Cell Biology. The molecular basis of human transcription-coupled DNA repair
Drugs That Target RNA Polymerase
Because bacterial RNA polymerase is structurally different enough from the human versions, it makes an excellent drug target. Rifampicin, the backbone antibiotic for tuberculosis treatment, works by binding tightly to the bacterial enzyme near its active center and physically blocking the growing RNA chain from extending past two or three nucleotides. It is remarkably potent, inhibiting half of E. coli RNA polymerase activity at a concentration of roughly 20 nanomolar.31PubMed Central. Inhibition of RNA Polymerase by Rifampicin and Rifamycin-Like Molecules
Rifampicin is not the only approach. Crystal structures of the bacterial enzyme bound to various antibiotics have revealed at least three distinct mechanisms of inhibition: blocking RNA extension (rifamycins and sorangicin), jamming the conformational cycle of the active center (streptolydigin), and preventing the enzyme from opening its cleft to load DNA in the first place (myxopyronin).32PubMed Central. Structures of RNA polymerase-antibiotic complexes Each targets a different vulnerability, which is why researchers continue to look for new binding sites as resistance to existing drugs grows.
Viral RNA Polymerases and What They Lack
RNA viruses face a problem that no cellular organism does: their genomes are made of RNA, so they need an enzyme that can copy RNA from an RNA template. These RNA-dependent RNA polymerases (RdRps) share some catalytic principles with cellular RNA polymerases, including a two-metal-ion mechanism and a conserved palm domain, but they are structurally simpler, typically acting as a single protein chain rather than a complex of a dozen or more subunits.33PubMed Central. RNA Dependent RNA Polymerases: Insights from Structure, Function and Evolution They also lack the elaborate proofreading, pausing, and regulatory features of their cellular counterparts. Coronaviruses are a notable exception; they encode a separate exonuclease that provides some error correction, which is one reason their genomes can be much larger than those of other RNA viruses.
The evolutionary relationship between viral and cellular RNA polymerases is an active area of research. Some large DNA viruses encode their own multisubunit RNA polymerases with clear homology to the bacterial enzyme. A distinct group of bacterial viruses (Caudovirales) carries deeply branching versions that may trace back to very early life, suggesting that viruses have played a role in the evolution of transcription machinery itself.34PubMed Central. A distinct lineage of Caudovirales that encodes a deeply branching multi-subunit RNA polymerase Whether these viral enzymes were ancestral to cellular ones or acquired from them remains debated, but the existence of such deeply branching forms complicates any simple narrative about where RNA polymerase came from.35PubMed. The origin of life: RNA and protein co-evolution on the ancient Earth