RNA Polymerase and the Transcription Cycle

RNA polymerase (RNAP) is the enzyme that catalyzes template-directed synthesis of RNA from DNA, the process called transcription. Unlike DNA polymerase, RNA polymerase can initiate a new chain de novo — it does not require a pre-existing primer — because its active site can catalyze formation of the first phosphodiester bond between two free ribonucleoside triphosphates positioned by the DNA template. The enzyme reads the template strand of DNA in the 3′→5′ direction and synthesizes a complementary RNA strand in the 5′→3′ direction, using the same Watson-Crick base-pairing rules as DNA replication, except that uracil (U) is incorporated opposite adenine (A) in place of thymine. All cellular RNA polymerases — bacterial, archaeal, and the three eukaryotic nuclear enzymes (Pol I, Pol II, Pol III) — share a conserved structural core built around two large subunits that form a "crab-claw" shape enclosing a central channel. This deep conservation, first appreciated when the bacterial and eukaryotic enzyme structures were solved and compared, tells us that the fundamental chemistry of transcription was established once, early in the history of life, and has been elaborated but never replaced. The transcription cycle itself is universally divided into three mechanistically distinct stages: initiation (finding the promoter, melting the DNA duplex, and beginning RNA synthesis), elongation (processively extending the RNA chain while translocating along the template), and termination (recognizing a stop signal and releasing both the RNA product and the DNA template). Each stage involves different conformations of the same enzyme, different accessory factors, and different kinetic bottlenecks — which is precisely why each stage is regulated independently and why each gets its own dedicated topic in the study of gene expression.

1. Enzyme architecture. The bacterial core RNA polymerase (found in organisms such as E. coli) has the subunit composition α₂ββ′ω, roughly 400 kDa in total.

• Two α subunits form a scaffold, help assemble the enzyme, and provide docking sites for transcription activator proteins.
• β and β′ are the two largest subunits and together form the crab-claw structure that clamps around DNA; they contribute the residues of the catalytic center, including the invariant aspartate triad that coordinates Mg²⁺.
• ω wraps around β′ and assists folding and assembly, and also has a chaperone-like role, stabilizing the enzyme under stress.

Core enzyme (α₂ββ′ω) can catalyze RNA synthesis but cannot recognize promoters specifically. Addition of a sigma (σ) factor converts core enzyme into holoenzyme, which can locate promoter sequences (foreshadowing the dedicated topics of Bacterial Promoters and the -10/-35 Elements and Sigma Factors and Promoter Recognition). Eukaryotic Pol II has an analogous but larger architecture — twelve subunits, with two large subunits (Rpb1, Rpb2) homologous to β′ and β — and requires an entire suite of general transcription factors (TFIIA–TFIIH) to substitute functionally for sigma.

2. Promoter binding and closed complex formation. Holoenzyme scans DNA largely by facilitated diffusion and sequence-nonspecific binding, then recognizes promoter elements through sigma. In bacteria, sigma-70 contacts two conserved sequence elements upstream of the transcription start site, typically centered around 10 and 35 base pairs upstream. Initial promoter binding produces a closed complex: the DNA duplex is still fully base-paired, but the enzyme is correctly positioned.

3. Open complex formation (DNA melting). The enzyme then unwinds roughly 13 base pairs of DNA around the start site, without any input of ATP hydrolysis — the energy comes from the favorable protein–DNA and protein–protein interactions formed as the enzyme clamps down, plus the AT-rich sequence typically found in the −10 element, which is intrinsically easier to melt. This produces the open complex (also called the transcription bubble), exposing the single-stranded template strand within the active-site cleft.

4. Initial (abortive) transcription. With the template strand positioned in the active site, the enzyme begins polymerizing RNA using two incoming NTPs positioned by base pairing with the template, joined by the catalytic center. Early in initiation the enzyme frequently releases short RNA transcripts of only a few to ~10 nucleotides and re-initiates at the same promoter — a wasteful-looking but mechanistically important phase called abortive initiation, thought to reflect the strain of holding both the upstream promoter contacts and the downstream growing RNA-DNA hybrid simultaneously.

5. Promoter escape. Once the nascent RNA reaches roughly 10–15 nucleotides, the enzyme undergoes a conformational change, breaks its contacts with the upstream promoter DNA and (in bacteria) typically releases sigma factor, and commits to processive elongation. This transition, promoter escape, marks the true beginning of the elongation phase.

6. Elongation. In the elongation complex, the enzyme maintains an approximately 8–9 base-pair RNA–DNA hybrid inside a transcription bubble of about 12–14 unwound base pairs, while re-annealing the DNA duplex behind it. For each nucleotide added:

• An incoming NTP diffuses into the active site and base-pairs with the template base positioned there.
• Catalysis proceeds by a two-metal-ion mechanism: Metal ion A, permanently chelated by the invariant aspartate triad, activates the RNA 3′-OH for nucleophilic attack on the α-phosphate of the incoming NTP; Metal ion B, delivered with the NTP, stabilizes the pyrophosphate leaving group. This chemistry is essentially the same used by DNA polymerases, replicative helicase-associated primases, and even some ribozymes — a striking case of catalytic convergence around Mg²⁺ ion pairs.
• The enzyme translocates one base pair downstream, a Brownian-ratchet-like motion aided by a mobile structural element called the bridge helix and a "trigger loop" that closes over the active site to enforce correct nucleotide selection, then swings open to permit translocation.
• The process repeats, typically adding nucleotides at 20–50 per second in bacteria (and roughly comparable but more variable rates in eukaryotic Pol II, further modulated by chromatin and pausing).

Elongation is not perfectly smooth — the enzyme pauses, can backtrack (sliding backward along the DNA so that the RNA 3′ end extrudes from the active site), and can proofread misincorporated nucleotides by cleaving the extruded RNA, topics developed further under Transcription Elongation and Proofreading.

7. Termination. Termination ends the cycle by destabilizing the RNA–DNA hybrid and the elongation complex, releasing RNA and freeing the enzyme to recycle. Bacteria use two principal mechanisms:

• Intrinsic (Rho-independent) termination: the DNA template encodes a GC-rich inverted repeat followed by a run of A's (making the nascent RNA's stretch a run of U's). The RNA folds into a stable hairpin immediately upstream of the RNA–DNA hybrid; this hairpin, combined with the intrinsically weak rU–dA base pairing, destabilizes the hybrid and causes the complex to fall apart — no additional protein required.
• Rho-dependent termination: the protein Rho, an ATP-dependent hexameric helicase, loads onto an untranslated, C-rich stretch of nascent RNA (a rut site), translocates along the RNA using ATP hydrolysis, catches up to a paused polymerase, and mechanically unwinds the RNA–DNA hybrid.

Eukaryotic Pol II termination is coupled to RNA 3′-end cleavage and polyadenylation signals and involves the exonuclease "torpedo" mechanism, a more elaborate story that builds on these same principles.

The modern picture of RNA polymerase rests on several landmark experimental programs.

Discovery and reconstitution of RNA synthesis (Weiss, Hurwitz, and Chamberlin, late 1950s–1960s). Samuel Weiss and Jerard Hurwitz independently identified an enzymatic activity in cell extracts that synthesized RNA using a DNA template and required all four ribonucleoside triphosphates. Building on this, Michael Chamberlin and Paul Berg (1962) purified DNA-dependent RNA polymerase from E. coli and showed, using a defined cell-free system, that the enzyme produced RNA whose base composition matched that predicted from the DNA template by Watson–Crick pairing rules (with U replacing T), and that RNA synthesis absolutely required a DNA template and all four NTPs, and was blocked by DNase treatment of the template. This established that RNA polymerase is a true template-directed, DNA-dependent enzyme, not simply a nucleotide-polymerizing enzyme acting independently of instruction — the crucial fact underlying the entire central dogma of transcription.

Discovery of sigma factor (Richard Burgess, Andrew Travers, Walter Zillig, and colleagues, 1969). By chromatographically fractionating purified E. coli RNA polymerase, Burgess and coworkers found that the enzyme separated into a "core" fraction (α₂ββ′) and an additional dissociable polypeptide they called sigma. Critically, core enzyme alone transcribed DNA nonspecifically — initiating at many random sites, including on both strands — while adding purified sigma back to core enzyme restored selective initiation at authentic promoter sites, verified by comparing the RNA products transcribed from bacteriophage DNA. This experiment proved that promoter specificity is a separable, modular property of the enzyme, conferred by a dissociable subunit rather than being an intrinsic, fixed property of the core catalytic machine — the founding observation behind the whole subject of Sigma Factors and Promoter Recognition.

Crystal structures of RNA polymerase (Roger Kornberg and coworkers, Seth Darst, and others, 1999–2001). Roger Kornberg's laboratory solved the first high-resolution crystal structure of yeast RNA polymerase II, visualizing the enzyme with a DNA template and nascent RNA transcript trapped in the act of transcribing (earning Kornberg the 2006 Nobel Prize in Chemistry). Around the same time, Seth Darst's group solved the structure of bacterial core RNA polymerase from Thermus aquaticus. These structures directly visualized the crab-claw architecture, the internal channel through which DNA threads, the bridge helix and trigger loop near the active site, and — using soaked-in heavy metal ions — located the catalytic Mg²⁺ ion coordinated by the invariant aspartate triad, confirming the two-metal-ion mechanism previously inferred biochemically for DNA polymerases. This structural work converted a biochemical black box into an atomically resolved machine, and revealed the deep structural homology between bacterial and eukaryotic polymerases predicted from decades of comparative biochemistry.

Single-molecule studies of elongation (Carlos Bustamante, Steven Block, and colleagues, 1990s–2000s). Using optical tweezers to attach a single RNA polymerase molecule to a bead and stretch the DNA template, these experiments measured the force generated by a single transcribing enzyme (found to be several piconewtons, comparable to some molecular motors), directly observed real-time pausing and backtracking events, and showed that RNAP moves in discrete steps corresponding to single-nucleotide addition. This provided direct mechanical evidence for the pause/backtrack/proofread behavior inferred from bulk biochemistry, foreshadowing the topic of Transcription Elongation and Proofreading.

Example 1 — Estimating a gene's transcription time and burst size. Suppose a bacterial operon is 4,500 base pairs long and RNA polymerase elongates at an average rate of 45 nucleotides per second in vivo (a typical rate, though pausing reduces the effective rate below the maximal chemical rate). Time to transcribe once: ``` 4,500 nt ÷ 45 nt/s = 100 seconds ``` Now suppose this operon has a strong promoter that permits reinitiation every 3 seconds (a new polymerase escapes the promoter every 3 s, a realistic value for a highly active rRNA-type promoter), and the gene stays "on" for a 10-minute (600 s) burst of activity. Number of polymerases that initiate during the burst: ``` 600 s ÷ 3 s per initiation = 200 polymerases ``` Because transcription of one 4,500-bp gene takes 100 s but new polymerases pile on every 3 s, many polymerases are simultaneously elongating along the same gene at once — up to roughly 100 s ÷ 3 s ≈ 33 polymerases in transit at any instant, densely loaded like cars on a highway. This illustrates a key point often missed by beginners: transcription of a single gene is not a solitary event but frequently a convoy of many RNA polymerase molecules moving in tandem, each independently completing initiation, elongation, and termination.

Example 2 — Diagnosing a rifampicin-resistant strain. A clinical microbiology lab isolates a strain of Mycobacterium tuberculosis that fails to respond to the antibiotic rifampicin, which works by binding within the DNA/RNA channel of the β subunit and blocking elongation of the nascent RNA once it reaches 2–3 nucleotides in length (it does not block initiation of the first phosphodiester bond, only the emergence of a longer RNA). Sequencing shows a point mutation in rpoB, the gene encoding the β subunit, within a small stretch known as the rifampicin resistance-determining region (RRDR). Reasoning through the mechanism:

• Rifampicin's target is the β subunit of RNA polymerase itself, not a peripheral accessory factor.
• A single amino-acid substitution in the rifampicin-binding pocket of β can sterically block drug binding while leaving the catalytic center and NTP channel intact, since the drug-binding surface and the catalytic aspartate triad are physically distinct regions of the enzyme.
• Because RNA polymerase is essential and single-copy per genome in bacteria (no redundant paralog to compensate), a mutation here is one of the most common mechanisms of clinically observed drug resistance — which is why rpoB sequencing is now a standard rapid molecular diagnostic (e.g., GeneXpert MTB/RIF) for tuberculosis drug resistance, allowing clinicians to switch therapy without waiting weeks for a culture-based test.

This case shows how a mechanistic model of the enzyme (which subunit, which pocket, which step is blocked) translates directly into a real diagnostic and treatment decision.

Roger David Kornberg came from a scientific family — his father Arthur Kornberg had won the Nobel Prize in 1959 for discovering DNA polymerase — and Roger turned to the sister enzyme, RNA polymerase, choosing the technically formidable challenge of solving its atomic structure. Working with yeast RNA polymerase II, a twelve-subunit, half-a-megadalton machine, his laboratory spent much of the 1990s developing crystallization methods robust enough to capture such a large, flexible, multi-subunit assembly, eventually trapping the enzyme mid-transcription with DNA template and nascent RNA visible in the electron density map. The resulting structures, published around 2001, showed with atomic clarity the internal channels for DNA entry, NTP entry, and RNA exit, and the mobile elements (bridge helix, trigger loop, clamp) that drive nucleotide addition and translocation — converting decades of biochemical inference about "the enzyme" into a literal, visualizable machine.

RNA polymerase is one of the most heavily exploited drug targets in medicine and biotechnology. Rifamycins (rifampicin, rifabutin) bind the bacterial β subunit and are frontline drugs against tuberculosis and leprosy; resistance-conferring mutations in rpoB are now tracked by rapid molecular diagnostics worldwide, directly shaping WHO tuberculosis treatment guidelines. The mushroom toxin α-amanitin binds and blocks eukaryotic RNA polymerase II specifically, a fact exploited both in poisoning diagnosis (distinguishing Amanita mushroom poisoning from other gastrointestinal illness) and in laboratory experiments to dissect which cellular processes depend on Pol II versus Pol I or Pol III. In biotechnology, bacteriophage RNA polymerases (especially T7 RNA polymerase, a small, single-subunit enzyme that recognizes a short, specific promoter with extremely high processivity) are workhorses of recombinant protein expression and of in vitro RNA synthesis, including the manufacture of mRNA vaccines, where T7 or SP6 polymerase is used to transcribe synthetic mRNA (such as the modified-nucleoside mRNA vaccines against SARS-CoV-2) from a linearized DNA template at industrial scale. Understanding promoter escape and processivity directly informs how such in vitro transcription reactions are optimized for yield and for minimizing double-stranded RNA byproducts that trigger unwanted innate immune responses. In agriculture and ecology, engineered promoters and polymerase-recognition elements are used to build inducible gene circuits in crops and in biosensor organisms, while comparative studies of RNA polymerase sequence and structure across bacteria, archaea, and eukaryotes remain a cornerstone of deep phylogenetic analysis, since the core subunits are among the most conserved proteins known, useful for reconstructing the tree of life at its deepest branches.

Ask the tutor: How does promoter strength (the exact −10/−35 sequence) quantitatively change the rate of open complex formation?