# Chapter 20. Bacterial RNA Polymerase: Structure, Catalysis, Initiation, and Elongation

## Scope Note

This chapter explains how bacterial DNA-dependent RNA polymerase reads promoter DNA, melts a transcription start region, begins RNA synthesis, enters processive elongation, pauses, proofreads, and terminates. The central object is bacterial multisubunit RNA polymerase, often abbreviated RNAP. RNAP is a molecular machine, an enzyme, a DNA-binding protein, an RNA-binding protein, and a regulatory hub at the same time. The chapter follows RNAP from core-enzyme assembly and sigma-factor-dependent promoter recognition through nucleotide addition, translocation, pausing, backtracking, intrinsic termination, Rho-dependent termination, antitermination, inhibitor action, and the experimental methods used to infer these states.

The chapter uses bacteria as the primary biological context, with recurring examples from Escherichia coli, alternative sigma-factor systems, sigma-54-dependent initiation, phage lambda antitermination, and antibiotic-sensitive bacterial RNAP states. Archaeal and eukaryotic polymerases are compared in [Chapter 21](chapter1020.md). Specialized organellar, single-subunit, and phage-like polymerases are treated in [Chapter 22](chapter1021.md). RNA-dependent RNA polymerases and reverse transcriptases are treated in [Chapter 23](chapter1022.md). The broader logic of elongation control, pausing, termination, and transcription-coupled processes continues in Chapters [24](chapter1023.md) and [25](chapter1024.md).

## Executive Summary

Bacterial RNA polymerase is a multisubunit enzyme that synthesizes RNA from a DNA template. The core enzyme contains the catalytic machinery for RNA-chain elongation, but efficient promoter-specific initiation usually requires a sigma factor. Sigma factors convert core RNAP into holoenzymes with defined promoter preferences. A housekeeping sigma factor directs RNAP to common growth-related promoters, while alternative sigma factors redirect transcription during stress, developmental transitions, stationary phase, envelope damage, heat shock, sporulation, or nitrogen limitation. Sigma factors are not passive labels attached to RNAP. They contact promoter DNA, help position duplex DNA, assist local DNA melting, influence start-site selection, and shape the probability that an initiation complex will escape the promoter.

Transcription initiation is a sequence of reversible and branching states. RNAP first forms closed promoter complexes in which promoter DNA remains mostly double-stranded. The enzyme then forms open complexes in which DNA near the transcription start site is melted and the template strand becomes available for base pairing with incoming ribonucleoside triphosphates. Initial RNA synthesis begins while RNAP still maintains promoter and sigma contacts, so the enzyme often makes and releases short abortive RNA products before producing a transcript long enough to support promoter escape. Abortive initiation is therefore a normal kinetic behavior of many promoters, not a sign that RNAP is broken.

During elongation, RNAP protects a transcription bubble and a short DNA-RNA hybrid while adding ribonucleotides to the RNA 3′ end. Each nucleotide addition cycle includes NTP selection, active-site alignment, phosphodiester-bond formation, pyrophosphate release, and translocation. The catalytic center uses divalent metal ions and conserved active-site residues to position the RNA 3′ hydroxyl and incoming NTP. Mobile structural elements, including the bridge helix, trigger loop, clamp, and secondary channel, help couple substrate selection, catalysis, and movement along DNA.

Elongation is processive but not uniform. RNAP can pause without falling off DNA. Pauses allow time for nascent RNA folding, regulatory factor recruitment, translation coupling, termination-factor action, and proofreading. Backtracking is a special paused state in which RNAP moves backward on the nucleic-acid scaffold and the RNA 3′ end leaves the active site. Backtracked complexes can be rescued by transcript cleavage, often stimulated by Gre factors in bacteria, which creates a new RNA 3′ end aligned for elongation. Misincorporation can increase pausing and favor correction pathways, so transcription fidelity depends on substrate selection, extension kinetics, backtracking, cleavage, and cellular context.

Termination releases the RNA transcript and DNA template from the elongation complex. Intrinsic termination uses a nascent RNA hairpin followed by a weak, often U-rich RNA-DNA hybrid to destabilize RNAP. Rho-dependent termination uses Rho, a hexameric RNA-dependent ATPase, which loads onto exposed nascent RNA and uses ATP-dependent movement to terminate a susceptible paused elongation complex. Antitermination systems alter RNAP or its associated factors so that termination signals are ignored or weakened. Phage lambda N antitermination is a classic example in which a nascent RNA element and host Nus factors form a processivity-modifying complex.

Bacterial RNAP is a validated antimicrobial target. Rifamycins and other inhibitor classes exploit bacterial-specific RNAP surfaces or transcription states, especially early initiation and elongation intermediates. RNAP inhibitors are also mechanistic probes because they freeze or bias particular states of the transcription cycle. Their therapeutic usefulness depends on target conservation, bacterial permeability, efflux, toxicity, resistance mutations, and the fitness cost of resistance.

The central lesson is that bacterial transcription is not controlled at one point. Regulation is distributed across promoter recognition, DNA opening, abortive initiation, promoter escape, elongation rate, pausing, proofreading, termination, antitermination, factor binding, nucleotide availability, DNA topology, translation coupling, and inhibitor-sensitive conformational states.

## Concept Inventory

- **Companion concept inventory defines the core vocabulary for this chapter. The most important terms:** introduced here because they are needed before the mechanisms can be followed.
- **Bacterial RNA polymerase core enzyme:** the multisubunit RNAP assembly capable of RNA elongation on a suitable template but lacking the sigma factor required for efficient promoter-specific initiation. In many bacteria the core enzyme is described as alpha2-beta-beta-prime-omega. The beta and beta-prime subunits form most of the catalytic cleft and nucleic-acid-binding channel, the alpha subunits support assembly and regulatory contacts, and omega stabilizes folding and assembly in many bacterial RNAPs.
- **RNA polymerase holoenzyme:** core RNAP bound to a sigma factor. Holoenzyme is the promoter-recognition form for most sigma-dependent initiation. The term should not be treated as one universal entity because different sigma factors produce different promoter preferences, open-complex lifetimes, and initiation kinetics.
- **Sigma factor:** a dissociable bacterial transcription specificity factor. A sigma factor recognizes promoter elements, helps position promoter DNA, assists DNA melting near the transcription start site, and influences promoter escape. Sigma-70-family factors and sigma-54-family factors use different initiation logic. Sigma-54-dependent initiation requires ATP-dependent activators and should not be treated as a simple variation of ordinary sigma-70 initiation.
- **Promoter:** a DNA region that directs transcription initiation. A promoter contains sequence information for RNAP recruitment, DNA opening, start-site selection, and regulation. In bacteria, common promoter features include -35 and -10 elements, extended -10 elements, discriminator regions, upstream elements, activator-binding sites, and initial-transcribed sequences, but not every promoter uses every feature.
- **Closed promoter complex:** an RNAP-promoter complex in which promoter DNA remains largely double-stranded. Open promoter complex, often abbreviated RPo, is a promoter-bound state in which DNA around the transcription start site is locally melted. Open-complex formation is necessary for template-strand access, but a melted promoter is not automatically a productive transcript-output event.
- **Abortive initiation:** repeated synthesis and release of short RNA products by promoter-bound RNAP before promoter escape. Abortive initiation can be a normal consequence of starting RNA synthesis while promoter contacts are still retained.
- **Promoter escape:** the transition from initially transcribing promoter-bound RNAP to a stable elongation complex. Escape requires rearrangement of promoter contacts, sigma contacts, DNA scrunching or related DNA movements, nascent RNA positioning, and RNAP conformational states.
- **Nucleotide addition cycle:** the repeated enzymatic cycle in which RNAP selects an incoming NTP, forms a phosphodiester bond at the RNA 3′ end, releases pyrophosphate, and translocates to expose the next template base.
- **Pausing:** a transient elongation state in which RNAP remains bound to DNA and RNA but nucleotide addition slows or stops. Pausing is not identical to termination or permanent arrest.
- **Backtracking:** reverse movement of RNAP along DNA and RNA that displaces the RNA 3′ end from the active site. Backtracking can enable proofreading when cleavage restores a properly aligned RNA 3′ end, but persistent backtracking can become arrest.
- **Intrinsic termination:** factor-independent bacterial termination driven by a nascent RNA hairpin and a weak RNA-DNA hybrid. Rho-dependent termination is termination mediated by the Rho RNA translocase, which loads on nascent RNA and uses ATP to act on paused transcription complexes. Antitermination is any regulatory mechanism that modifies the elongation complex or nascent RNA environment so RNAP reads through termination signals.

## What to Know Before Reading This Chapter

The reader should know three basic transcription facts. First, RNA is synthesized 5′ to 3′, which means each incoming ribonucleoside triphosphate is added to the 3′ hydroxyl of the growing RNA. Second, the DNA template strand is read 3′ to 5′, so the RNA sequence is complementary to the template strand and similar to the non-template strand except that RNA contains uracil instead of thymine. Third, a bacterial gene annotation such as promoter or terminator is a genome-level label for a biochemical behavior. [Chapter 15](chapter1014.md) introduces promoters, transcription units, operons, and terminators as genomic objects. This chapter asks what RNAP physically does with those objects.

The reader should also separate three levels of explanation. A sequence motif is a pattern in DNA or RNA. A biochemical intermediate is a molecular state such as a closed complex, open complex, paused complex, or backtracked complex. A cellular outcome is a change in transcript abundance, transcript boundary, gene expression, stress survival, or antibiotic sensitivity. The same motif can behave differently in different cellular contexts because the outcome depends on factor concentration, DNA supercoiling, NTP levels, translation, RNA folding, and competing kinetic pathways.

Four running examples recur through the chapter. An Escherichia coli sigma-70 promoter illustrates ordinary sigma-dependent initiation. Sigma-54-dependent promoters illustrate an alternative initiation system requiring ATP-dependent activators. A hairpin-dependent pause or terminator illustrates how nascent RNA structure can feed back on RNAP. Phage lambda N antitermination illustrates how a nascent RNA element and protein factors can convert RNAP into a termination-resistant elongation complex.

## 20.1. Core enzyme, sigma factors, and promoter recognition

Bacterial RNAP core enzyme is the catalytic platform of bacterial transcription. It contains a deep cleft that binds nucleic acids, an active site for phosphodiester-bond formation, channels for incoming nucleotides and RNA exit, and mobile elements that change position during initiation and elongation. The core enzyme can synthesize RNA when it is placed on an appropriate nucleic-acid scaffold, which is why purified core RNAP can be studied in elongation assays. In cells, however, the problem is not only catalysis. RNAP must find the right promoter, melt DNA at the right position, choose a transcription start site, and avoid initiating from random DNA. Sigma factors solve much of this specificity problem by forming holoenzymes with promoter-recognition capacity.

![Figure 20.1. Bacterial Transcription-Cycle State Map](../assets/figures/chapter1019_figure1.png)

**Figure 20.1. Bacterial Transcription-Cycle State Map.** Bacterial RNA polymerase moves through a series of branching states rather than a single linear path. Starting from the closed promoter complex, the enzyme proceeds to an open complex in which promoter DNA is locally melted, then enters an initially transcribing complex that can repeatedly release short abortive RNAs before achieving promoter escape. During elongation, RNAP can pause, backtrack with displacement of the RNA 3′ end from the active site, or terminate; backtracked complexes can be rescued by transcript cleavage and restored to productive elongation.

![Figure 20.2. Structural Organization of Bacterial RNAP Holoenzyme](../assets/figures/chapter1019_figure2.png)

**Figure 20.2. Structural Organization of Bacterial RNAP Holoenzyme.** Bacterial RNAP holoenzyme is assembled from the core subunits α₂ββ′ω and a dissociable sigma factor that confers promoter specificity. The β and β′ subunits form the catalytic cleft and nucleic-acid-binding channel, while sigma contacts −35 and −10 promoter elements and must rearrange during early RNA synthesis. Mobile elements including the bridge helix, trigger loop, clamp, secondary channel, and RNA-exit channel each play distinct roles in substrate selection, catalysis, and translocation.

**Table 20.1. Bacterial RNAP Subunits and Mobile Elements.** The table summarizes the physical location, primary function, main evidence type, and a common misconception for each subunit and mobile element of bacterial RNA polymerase holoenzyme.

| Component | Physical Location | Primary Function | Evidence Type | Common Misconception |
| --- | --- | --- | --- | --- |
| **α (two copies)** | Flanking the catalytic cleft; C-terminal domains project outward | Assembly scaffold; C-terminal domains contact upstream DNA and activators | X-ray crystallography; biochemistry | Often treated as structural-only; CTD contacts regulate many promoters |
| **β** | Forms one jaw of the catalytic cleft | Contains most active-site residues; lines the nascent-RNA channel | X-ray crystallography; mutagenesis | Not a passive scaffold; many inhibitors and resistance mutations map here |
| **β′** | Forms the opposing jaw of the catalytic cleft | Completes the active site; grips DNA in the cleft | X-ray crystallography; mutagenesis | Not interchangeable with β; the two subunits divide active-site labour |
| **ω** | Wraps around the β′ C-terminus | Stabilizes β′ folding and core-enzyme assembly | Biochemistry; structural data | Often omitted from diagrams; loss can reduce RNAP levels in cells |
| **σ (sigma)** | Contacts the outer face of RNAP and promoter DNA | Promoter recognition; assists DNA melting; influences promoter escape | Biochemistry; structural; genetics | Not a permanent subunit; sigma release timing is context-dependent |
| **Bridge helix** | Spans the cleft between β and β′ near the active site | Influences translocation by linking catalytic and clamp motions | X-ray crystallography | Not a rigid strut; conformational flexibility is functionally important |
| **Trigger loop** | Folds into the active site upon correct NTP binding | Promotes catalysis and substrate discrimination | X-ray crystallography; kinetics | Not a simple gate; coupled to fidelity and elongation rate |
| **Clamp** | Movable domain of β′ above the cleft | Grips the nucleic-acid scaffold during elongation | Cryo-EM; crosslinking | Not locked shut; clamp opening allows nucleic-acid loading and is targeted by some inhibitors |
| **Secondary channel** | Narrow tunnel leading into the active site | Admits incoming NTPs; provides access for Gre factors and some inhibitors | Structural; biochemistry | Not the main DNA or RNA exit channel; restricted geometry shapes regulatory use |
| **RNA-exit channel** | Surface groove from which nascent RNA emerges | Guides RNA out of RNAP; interacts with hairpins involved in pausing and termination | Structural; crosslinking | Not a passive tube; contacts with hairpins affect pausing, termination, and factor binding |

The canonical bacterial core enzyme is written alpha2-beta-beta-prime-omega. This notation is more than a list of parts. The beta and beta-prime subunits form the jaws of the catalytic cleft and hold conserved active-site elements. The two alpha subunits participate in assembly and provide C-terminal domains that can contact upstream DNA elements or transcription activators. The omega subunit is small, but it can stabilize RNAP folding and assembly. A useful mental picture is a clamp-like enzyme that closes around a DNA-RNA scaffold while leaving channels for downstream DNA entry, upstream DNA exit, nascent RNA exit, incoming NTP access, and regulatory factor binding.

Sigma factors turn this catalytic platform into promoter-specific holoenzymes. A sigma factor is not a permanent subunit of all RNAP states. It binds core enzyme, recognizes promoter elements, assists DNA opening, and then rearranges during early RNA synthesis. In the sigma-70 family, domains commonly recognize -35 and -10 promoter elements and help stabilize the melted non-template strand around the start site. The exact promoter grammar varies. Some promoters have strong -35 and -10 elements. Others lack one canonical element but use an extended -10 region, upstream element, activator, or favorable discriminator sequence. Therefore, a promoter cannot be judged only by how closely it matches one consensus logo.

The bacterial cell uses multiple sigma factors to reprogram transcription. A housekeeping sigma factor directs expression of many genes needed during ordinary growth. Alternative sigma factors redirect RNAP to promoters needed in specific physiological contexts, such as heat shock, envelope stress, oxidative stress, motility, stationary phase, sporulation, or nitrogen starvation. This logic lets one core enzyme support many transcriptional programs. Competition among sigma factors can also make transcription sensitive to sigma abundance, anti-sigma factors, anti-anti-sigma factors, stress signaling, and RNAP availability.

**Table 20.2. Sigma-Factor Logic.** The table compares major sigma-factor classes by promoter features, regulatory context, initiation mechanism, and important boundary cases.

| Sigma Class | Promoter Features | Regulatory Context | Initiation Mechanism Notes | Boundary Cases |
| --- | --- | --- | --- | --- |
| **Housekeeping σ70 family** | −35 and −10 elements; optional upstream element and extended −10 | Constitutive and growth-related genes | Contacts both promoter elements; assists local DNA melting; rearranges at promoter escape | Some strong promoters lack a canonical −35 element; upstream element or extended −10 can compensate |
| **Heat-shock σ (e.g., σ32 in E. coli)** | Distinct −35 and −10 consensus recognized by σ32 | Protein-folding and damage-response genes during heat shock | Same open-complex logic as σ70 family but with a different sequence preference | σ32 is rapidly degraded after stress; competition with σ70 shapes response amplitude |
| **Envelope-stress σ (e.g., σE)** | σE-specific −35 and −10 elements | Periplasmic stress and membrane-integrity responses | Anti-sigma sequestration controls availability | σE regulation involves a proteolytic cascade; not a simple on/off switch |
| **Stationary-phase σS (σ38)** | Overlapping sequence specificity with σ70; minor differences | Stress survival, stationary phase, general stress response | Competes directly with σ70 for core RNAP | Distinguishing σ70 from σS promoters in vivo requires more than sequence inspection alone |
| **Sigma-54 (σ54 / σN)** | −24 and −12 elements (not −35/−10); requires enhancer-bound activator | Nitrogen limitation, flagella biosynthesis, and other adaptive responses | Closed complex forms normally, but open complex requires ATP hydrolysis by a bacterial enhancer-binding protein | Mechanistically distinct from all σ70-family factors; not a simple sequence variant of the same mechanism |

Sigma-54 is an important boundary case. Sigma-54-dependent promoters do not behave like ordinary sigma-70-family promoters with a different sequence logo. Sigma-54 holoenzyme can bind promoter DNA in a closed complex but requires an ATP-dependent activator, often bound at an enhancer-like site, to remodel the complex into an open state. This creates a regulated checkpoint in which promoter-bound RNAP is poised but inactive until an activator provides energy and conformational change.

Promoter recognition is best understood as a kinetic pathway. RNAP holoenzyme first encounters promoter DNA, samples sequence and shape, and forms a closed complex. If the contacts are favorable, RNAP bends and opens the DNA to form an open complex. Open-complex stability, start-site positioning, and promoter escape then determine how often productive transcripts are made. A promoter with tight RNAP binding may still be weak if it rarely opens or escapes. A promoter with modest binding can be strong if it rapidly forms productive open complexes. The boxed kinetic-composite discussion develops this point because it prevents a common misconception: promoter strength is not the same as binding affinity.

Structural and biochemical studies established that sigma occupies RNAP surfaces near DNA-entry and RNA-exit paths and must be rearranged as transcription begins. These studies also illustrate a broader evidence principle. A binding or scattering experiment can show factor association and overall arrangement, while a high-resolution structure can show molecular contacts, and a kinetic assay can show which contacts affect pathway rates. No single method fully defines promoter recognition.

Concrete example: many Escherichia coli sigma-70 promoters contain recognizable -35 and -10 elements separated by an appropriate spacer. Mutations that improve these elements often increase initiation, but not always. A change that improves binding can trap RNAP in a long-lived open complex or increase abortive initiation without increasing productive elongation. Conversely, upstream activator contacts can strengthen a promoter that has a weak core element. The correct question is not whether RNAP binds, but which step in the initiation pathway is favored or disfavored.

## 20.2. Initiation complexes, abortive initiation, and promoter escape

Initiation begins before the first phosphodiester bond is formed. A closed promoter complex contains RNAP holoenzyme bound to promoter DNA that remains mostly double-stranded. Closed complexes are often short-lived because RNAP samples many DNA sequences that are not productive promoters. At a true promoter, contacts with sigma, core RNAP, upstream DNA, and sometimes activators increase the probability that the complex will proceed to DNA opening.

Open-complex formation melts a short region of promoter DNA around the transcription start site. Melting means that the two DNA strands separate locally, not that the entire promoter becomes single-stranded. The template strand enters the active-site cleft so an incoming NTP can base-pair with the first template base. The non-template strand is stabilized by protein contacts and displaced away from the template strand. DNA supercoiling can influence this step because negatively supercoiled DNA is generally easier to open, although promoter-specific details matter. Footprinting and permanganate probing are classic ways to infer promoter opening: footprinting reveals protein-protected DNA, while permanganate preferentially modifies exposed thymines in single-stranded DNA. These assays report DNA accessibility, not complete catalytic competence.

The first few RNA bonds are made while RNAP still grips promoter DNA. This creates a mechanical problem. The enzyme must synthesize RNA downstream from the start site, but promoter contacts still hold the enzyme near its original position. One solution is DNA scrunching, in which downstream DNA is pulled into the enzyme while upstream promoter contacts remain. As the short RNA grows, strain accumulates in the initially transcribing complex. The complex can relieve strain by releasing a short abortive RNA and trying again, or by breaking promoter contacts and escaping into elongation.

Abortive initiation produces short RNAs, often only a few nucleotides long, that are released while RNAP remains promoter-bound. Abortive products are frequently detected in purified transcription assays as ladders of short RNAs. These RNAs are not usually functional transcripts. They are molecular records of initiation attempts. A promoter that produces many abortive RNAs is not necessarily defective; abortive initiation can be a natural kinetic cost of starting RNA synthesis from a promoter-bound state.

Promoter escape is the transition from initiation to elongation. The nascent RNA reaches a length at which it competes with sigma and promoter contacts, the DNA-RNA hybrid stabilizes a moving transcription bubble, and RNAP releases or rearranges interactions that held it at the promoter. Escape does not always mean immediate sigma loss. Sigma can dissociate, remain loosely associated, or influence early elongation at some promoters. This boundary case matters because the phrase sigma release is often used too casually. Promoter escape is defined by stable movement into elongation, not by a universal sigma-dissociation event.

![Figure 20.5. Promoter Contacts, DNA Opening, Scrunching, and Promoter Escape](../assets/figures/chapter1019_figure5.png)

**Figure 20.5. Promoter Contacts, DNA Opening, Scrunching, and Promoter Escape.** A matched-state structural sequence follows a canonical sigma-70-family holoenzyme from the closed complex through local promoter melting, the initially transcribing complex, abortive release, and productive escape. Promoter DNA and RNAP are registered to the same reference position in the first three states so sigma-promoter contacts, transcription-bubble formation, downstream-DNA scrunching, short-RNA growth, and accumulated strain can be compared directly. Abortive RNA release returns the promoter-bound complex to another attempt, whereas productive growth rearranges promoter and sigma contacts and stabilizes downstream movement; sigma retention and dissociation are shown as alternative outcomes rather than a mandatory escape step.

Initiation kinetics determine promoter output. Some promoters form open complexes rapidly but escape slowly. Others open slowly but escape efficiently once open. Some are sensitive to the concentration of initiating NTPs because the first few bases require particular nucleotides. Some are strongly regulated by discriminator sequences, ppGpp, DksA, activators, repressors, or DNA topology. In bacterial physiology, these differences allow promoter classes to respond differently to nutrient limitation, stress, growth rate, and regulatory proteins.

Evidence for initiation pathways comes from complementary methods. Abortive transcript assays report short RNA products and initiation cycling. Promoter-mutagenesis experiments reveal sequence features that affect binding, opening, start-site choice, and escape. Footprinting reports DNA protection. Permanganate probing reports melted thymine exposure. Single-molecule fluorescence can distinguish long-lived and short-lived complexes. Cryo-EM and crystallography can reveal trapped initiation states. Kinetic modeling connects these observations into rate constants and branching probabilities. The limitation is that each method enriches or observes particular states. A structure can show a plausible state, but it does not by itself prove how fast the state forms or whether it dominates in cells.

Do not overgeneralize: a strong promoter is not one biochemical property. It is a composite of RNAP recruitment, closed-complex lifetime, open-complex formation, open-complex lifetime, initiating nucleotide availability, abortive-initiation probability, promoter-escape probability, and early elongation survival. The boxed kinetic-composite summary reinforces this point.

> **Box 20.1. Promoter Strength Is a Kinetic Composite**
>
> - Promoter output is not determined by a single binding event or sequence score.
> - It reflects the combined contribution of RNAP recruitment, closed-complex formation rate, open-complex formation rate, open-complex lifetime, initiating NTP availability, abortive-initiation frequency, promoter-escape probability, and early-elongation survival.
> - A near-consensus promoter can be weak if it forms a long-lived but non-productive open complex or rarely escapes into elongation.
> - A sub-optimal promoter can be strong if an activator, upstream element, or favorable discriminator accelerates a limiting step.
> - Regulatory proteins and signals typically act by changing one or more steps in this kinetic series, not by globally toggling RNAP binding on or off.

## 20.3. Active-site chemistry, nucleotide addition, and translocation

Once RNAP has escaped the promoter, it becomes a processive elongation complex. Processive means that RNAP can add many nucleotides without dissociating after each addition. The elongation complex contains downstream double-stranded DNA, an opened transcription bubble, an RNA-DNA hybrid, upstream re-annealed DNA, and nascent RNA emerging from the RNA-exit channel. The RNA-DNA hybrid is short enough to fit within the active-site cleft but long enough to stabilize the complex. In many textbook diagrams the hybrid is drawn as a static ladder. In reality, it is a dynamic part of a moving enzyme.

The nucleotide addition cycle begins when an incoming ribonucleoside triphosphate enters the active site through the secondary channel and pairs with the next template base. The correct NTP must be selected from ATP, GTP, CTP, and UTP, and cellular NTP concentrations can influence elongation kinetics. The RNA 3′ hydroxyl attacks the alpha phosphate of the incoming NTP, forming a new phosphodiester bond and releasing pyrophosphate. The reaction depends on divalent metal ions and conserved active-site residues that position substrates and stabilize charge during catalysis. After chemistry, RNAP must translocate by one base pair so the newly extended RNA 3′ end moves away from the addition site and the next template base becomes available.

Translocation is not a separate motor stroke like the stepping of some cytoskeletal motors. RNAP can fluctuate thermally between pre-translocated and post-translocated states, and incoming NTP binding can stabilize the post-translocated state. Sequence-dependent hybrid stability, downstream DNA contacts, RNA structures, elongation factors, and active-site conformational changes influence this equilibrium. The bridge helix and trigger loop are central mobile elements in this discussion. The bridge helix lies near the active site and is positioned to influence translocation. The trigger loop can fold around the incoming NTP and promote catalysis. These conformational changes help explain why substrate selection, catalysis, and movement are coupled rather than independent events.

Fidelity in transcription means that RNAP usually incorporates the ribonucleotide complementary to the DNA template. Fidelity arises at several stages. First, correct base pairing and active-site geometry favor the right NTP before chemistry. Second, incorrect NTPs are more likely to dissociate before incorporation. Third, a mismatched RNA 3′ end can slow subsequent extension. Fourth, the elongation complex can backtrack, allowing cleavage of the misincorporated nucleotide from the RNA. These layers make RNAP more accurate than a simple passive copying machine. At the same time, RNAP fidelity is not as high as DNA-replication fidelity because RNA errors usually affect individual RNA molecules rather than permanently altering the genome.

Misincorporation can produce pauses. If the wrong nucleotide is incorporated, the resulting mismatch can distort the hybrid or active site, making the next addition slow. This slowing increases the chance of backtracking and correction. Biochemical and kinetic studies support the idea that misincorporation is a major source of transcriptional pausing and can feed into fidelity mechanisms. This is a useful conceptual link: pausing is not only regulatory; it can also be a biochemical consequence of an error.

RNA synthesis also has a reverse reaction, pyrophosphorolysis, in which pyrophosphate contributes to removal of the terminal nucleotide under some conditions. In cells, transcript cleavage stimulated by factors is usually more prominent in proofreading discussions, but pyrophosphorolysis is important for understanding the chemical reversibility of polymerase reactions and some inhibitor mechanisms.

The active site is reachable through the secondary channel, which also admits regulatory factors and some inhibitors. Gre factors can enter through this channel and stimulate cleavage of backtracked RNA. Secondary-channel inhibitors can interfere with nucleotide access or active-site function. This explains why a narrow structural route into the active site can have broad regulatory and pharmacological significance.

Do not overgeneralize: the nucleotide addition cycle is often drawn as a neat loop, but real elongation is interrupted by pauses, factor binding, backtracking, NTP limitation, RNA folding, DNA-bound proteins, supercoiling, DNA damage, and collisions with replication or translation machinery. A simple loop is a useful entry diagram, not a complete model of cellular transcription.

![Figure 20.3. Nucleotide Addition, Pausing, and Backtracking](../assets/figures/chapter1019_figure3.png)

**Figure 20.3. Nucleotide Addition, Pausing, and Backtracking.** Each nucleotide addition cycle involves NTP selection, phosphodiester-bond formation, pyrophosphate release, and translocation; however, elongation is not uniform. RNAP can enter an elemental or hairpin-stabilized paused state, or backtrack so that the RNA 3′ end is displaced from the active site into the secondary channel; in the latter case, Gre-factor-stimulated cleavage of the extruded RNA restores a correctly positioned 3′ end and allows elongation to resume.

## 20.4. Pausing, backtracking, proofreading, and fidelity

Pausing is a temporary state in which RNAP remains bound to the transcription scaffold but slows or stops nucleotide addition. A pause is different from termination because the enzyme has not released RNA and DNA. A pause is different from permanent arrest because the enzyme can resume elongation. Pauses can last milliseconds, seconds, or longer depending on sequence, factors, NTP concentrations, and cellular conditions. Many short pauses may be incidental consequences of local sequence and enzyme dynamics. Some pauses are regulatory because they affect RNA folding, factor recruitment, termination, translation coupling, or gene-expression timing.

Several mechanisms can create or stabilize pauses. An elemental pause can arise from active-site or translocation-state changes without a large RNA structure. A hairpin-stabilized pause occurs when a nascent RNA hairpin forms near the RNA-exit channel and interacts with RNAP or associated factors. A backtracked pause occurs when RNAP moves upstream relative to DNA and RNA, extruding the RNA 3′ end away from the active site. A regulatory pause can combine these features. For example, a nascent RNA hairpin may stabilize a paused state long enough for a ribosome, attenuation system, Rho factor, or RNA-binding protein to act.

Backtracking is especially important because it links pausing, proofreading, and arrest rescue. In a backtracked complex, the RNA 3′ end is no longer aligned for NTP addition. The extruded RNA can occupy the secondary channel. RNAP can recover by diffusing forward, but deeper backtracking often requires transcript cleavage. GreA and GreB in bacteria stimulate cleavage of the extruded RNA, generating a new RNA 3′ end in the active site. This restores elongation competence and can remove misincorporated residues.

Pausing is also central to bacterial gene regulation. In attenuation-like systems, a pause can provide time for a leader RNA to fold or for a translating ribosome to reach a decision point. In riboswitch systems, transcriptional pausing can give a metabolite-binding aptamer time to bind ligand before the expression platform folds into a terminator or antiterminator. In Rho-dependent termination, a pause can give Rho time to catch RNAP. In co-transcriptional RNA folding, pauses influence which RNA structures form first. These topics are developed in Chapters [24](chapter1023.md), [25](chapter1024.md), [78](chapter1073.md), and [79](chapter1074.md), but the enzymatic foundation is here: RNAP speed and pause lifetime can change RNA fate.

Nus factors illustrate that elongation factors do not have one universal effect. NusG-family proteins can increase elongation processivity, suppress some pauses, promote other pause or termination outcomes, and connect transcription to translation or termination machinery depending on system. In Bacillus subtilis, NusG can influence hairpin-dependent pausing in the trp leader. In other contexts, NusG-family factors participate in Rho termination, antitermination, or elongation acceleration. Therefore, a factor name alone is not enough; the relevant complex, RNA sequence, organism, and assay must be stated.

Fidelity and proofreading should be understood quantitatively rather than morally. RNAP sometimes makes errors. The question is how often, under what conditions, and with what consequences. An mRNA containing one wrong nucleotide may produce a mutant protein from that molecule, but the error is not inherited. Errors in rRNA, tRNA, regulatory RNAs, or highly expressed transcripts can still matter, especially if they affect folding or function. The evolutionary pressure on RNAP fidelity is therefore real but different from the pressure on DNA polymerase fidelity. The boxed fidelity comparison develops this point.

Evidence for pausing and backtracking comes from multiple scales. Purified transcription assays can map pause sites and measure lifetimes under defined conditions. Nucleotide-walk experiments can synchronize RNAP at a template position. Single-molecule force experiments can observe movement and backtracking. Nascent RNA sequencing can map RNAP positions in cells, although interpretation depends on library preparation, resolution, and whether the method captures paused, active, or processed complexes. Structural studies can show paused or backtracked conformations, but structures are often stabilized by scaffold design, mutations, factors, or inhibitors. A pause observed in vitro is a hypothesis for cellular regulation, not proof by itself.

Do not overgeneralize: RNAP paused here is not equivalent to this pause regulates the gene. Regulatory interpretation requires evidence that the pause occurs in the relevant cell type or condition, has enough lifetime to affect a downstream event, changes when the proposed regulatory factor or signal is perturbed, and alters transcript output, RNA folding, translation coupling, termination, or fitness.

> **Box 20.2. Pausing Is Not Automatically Regulation**
>
> - Many RNAP pause sites are detected in vitro or by nascent-RNA sequencing, but most are incidental consequences of local sequence and enzyme dynamics rather than dedicated regulatory events.
> - A pause should be called regulatory only when four criteria are met: it occurs in the relevant cell type or condition; it has sufficient lifetime to affect a downstream process; it changes when the proposed regulatory signal or factor is perturbed; and it alters transcript output, RNA folding, translation coupling, termination, or fitness.
> - Failing to distinguish incidental from regulatory pauses leads to overclaiming about the extent of transcriptional control.

> **Box 20.3. Why RNA Polymerase Fidelity Differs from DNA Polymerase Fidelity**
>
> - Transcription errors usually affect only individual RNA molecules, not the heritable genome, so the evolutionary cost of a single transcription error is lower than the cost of a replication error.
> - RNAP fidelity arises at multiple stages: NTP selection and active-site geometry before chemistry, slow extension of a mismatched RNA 3′ end, backtracking, and Gre-stimulated transcript cleavage.
> - Despite lower proofreading pressure than DNA replication, RNAP fidelity still matters for rRNA folding, tRNA function, highly expressed mRNAs, and regulatory RNAs where errors can impair function.
> - This comparison clarifies why misincorporation rates for RNAP are orders of magnitude higher than for high-fidelity DNA polymerases, yet transcription remains biologically adequate.

## 20.5. Intrinsic termination, Rho-dependent termination, and antitermination

Termination is the controlled release of the RNA transcript and disengagement of RNAP from the DNA template. Because elongation complexes are highly stable, termination requires specific mechanisms that destabilize the complex at the right position. Bacteria use two major termination classes: intrinsic termination and Rho-dependent termination. They also use antitermination systems that allow RNAP to ignore termination signals when continued transcription is needed.

![Figure 20.4. Intrinsic and Rho-Dependent Termination as Competing Kinetic Pathways](../assets/figures/chapter1019_figure4.png)

**Figure 20.4. Intrinsic and Rho-Dependent Termination as Competing Kinetic Pathways.** Intrinsic termination is driven by a GC-rich nascent RNA hairpin followed by a U-rich RNA–DNA hybrid that destabilizes the elongation complex without requiring any protein factor beyond RNAP itself. Rho-dependent termination depends on Rho, a hexameric ATP-dependent RNA translocase, which loads onto exposed C-rich nascent RNA and uses ATP hydrolysis to translocate and terminate a susceptible paused elongation complex. Both pathways are kinetically sensitive to RNAP elongation rate, RNA folding, ribosome occupancy, and factor availability.

Intrinsic termination, also called factor-independent termination or rho-independent termination, usually requires two RNA features: a GC-rich hairpin followed by a U-rich tract in the RNA. The hairpin forms in the nascent RNA as it exits RNAP. The U-rich tract pairs weakly with the DNA template as rU-dA base pairs in the RNA-DNA hybrid. A paused RNAP at this sequence gives the hairpin time to form. The hairpin and weak hybrid together destabilize the elongation complex, leading to transcript release.

Intrinsic terminator efficiency is not determined only by the presence of a hairpin and U tract. It depends on hairpin folding kinetics, hairpin stability, hybrid stability, pause lifetime, elongation rate, ionic conditions, transcription factors, and upstream RNA structures. A terminator hairpin that forms too slowly may fail if RNAP moves past the termination window. A U-rich tract may be insufficient if the hairpin is weak. A factor that changes RNAP pausing can change termination without binding the terminator RNA directly. This is why terminators should be treated as kinetic devices, not just sequence motifs.

Rho-dependent termination uses a protein factor. Rho is a hexameric ATP-dependent RNA translocase. It loads onto exposed nascent RNA regions often called rut sites. These regions are commonly C-rich and relatively unstructured, but the precise rules vary. After loading, Rho uses ATP hydrolysis to move along RNA. Termination occurs when Rho catches a susceptible RNAP, often one that is paused, and promotes release of RNA and DNA.

Translation strongly affects Rho access in bacteria. When ribosomes closely follow RNAP on a protein-coding gene, the nascent RNA is occupied and Rho-loading sites may be hidden. If translation is inefficient, premature stop codons appear, mRNA is poorly engaged by ribosomes, or untranslated RNA is exposed, Rho can terminate transcription. This creates transcriptional polarity: a defect early in an operon can reduce expression of downstream genes because Rho terminates before RNAP reaches them. Rho also suppresses pervasive or antisense transcription and contributes to genome defense by terminating transcripts that should not continue. Do not overgeneralize Rho as a factor that only removes aberrant transcripts; Rho also participates in normal regulation.

Antitermination is the opposite regulatory logic. Instead of causing release, antitermination modifies RNAP or the nascent RNA environment so the elongation complex reads through terminators. Phage lambda N antitermination is a classic example. The N protein binds a nascent RNA element and recruits host Nus factors to form an RNP complex that modifies RNAP processivity and termination response. This allows phage transcription to continue through terminators that would otherwise stop RNAP. The example is powerful because it shows how an RNA element made by RNAP can feed back onto the same RNAP molecule.

Antitermination also occurs in bacterial ribosomal RNA operons and regulatory leader regions. These systems can increase processivity, alter pausing, prevent terminator hairpin action, block Rho access, or recruit factors that change RNAP behavior. The shared principle is that termination is a competition. Elongation, pausing, RNA folding, Rho loading, hairpin formation, factor binding, and ribosome movement all occur on overlapping timescales. A small change in one rate can change whether a transcript is released or extended.

Evidence for termination mechanisms includes terminator mutagenesis, in vitro transcription with purified RNAP and factors, RNA-structure probing, Rho ATPase and RNA-binding assays, nascent RNA mapping, genetic suppression, and structural studies of termination or antitermination complexes. Each evidence type has limits. A strong terminator in vitro may be weaker in vivo if a ribosome prevents hairpin formation or a factor suppresses pausing. A predicted rut site may not function if the RNA folds or is protein-bound. A readthrough transcript can reflect antitermination, terminator mutation, faster elongation, factor depletion, or altered RNA stability. Good interpretation requires connecting transcript boundaries to mechanism.

**Table 20.3. Termination and Antitermination Mechanisms.** The table compares bacterial termination and antitermination mechanisms by required RNA features, protein factors, kinetic requirements, and representative examples.

| Mechanism | Required RNA Feature | Required Protein Factor | Main Kinetic Requirement | Example | Cross-Reference |
| --- | --- | --- | --- | --- | --- |
| **Intrinsic termination** | GC-rich hairpin followed by U-rich RNA–DNA hybrid | None (factor-independent) | Hairpin must fold before RNAP moves past the termination window; pausing increases efficiency | Many bacterial gene and operon terminators in E. coli | Ch. 23 |
| **Rho-dependent termination** | Exposed, relatively unstructured C-rich nascent RNA (rut site) | Rho hexameric ATPase | Rho must load and translocate to catch a paused RNAP before elongation resumes | Transcriptional polarity; termination in untranslated or poorly translated regions | Ch. 23 |
| **Lambda N antitermination** | boxB RNA hairpin in the nascent transcript | N protein; host Nus factors (NusA, NusB, NusE, NusG) | Antitermination RNP must assemble on RNAP before it reaches downstream terminators | Phage λ early-gene expression reads through tL1 and tR1 terminators | Ch. 24 |
| **rRNA operon antitermination** | Nascent RNA leader elements (boxA, boxB, boxC) | NusB, NusE, NusA, NusG | High-processivity complex must assemble early to suppress intrinsic terminators throughout long rRNA operons | E. coli rrn operons | Ch. 24 |
| **Attenuation-like leader control** | Leader RNA capable of forming alternative hairpins (terminator vs. antiterminator) | Ribosome (and, in some cases, RNA-binding proteins) | Pause provides time for ribosome or ligand to steer RNA folding; outcome determines termination or readthrough | E. coli trp leader; B. subtilis trp attenuation | Ch. 73 |

## 20.6. Inhibitors, structural methods, kinetics, and regulation

Bacterial RNAP is one of the most important antibiotic targets in RNA biology. An inhibitor can block transcription by preventing initiation, early RNA extension, promoter escape, nucleotide addition, translocation, factor binding, or elongation-complex stability. Rifamycins are the best-known RNAP inhibitors. They bind bacterial RNAP near the path of the nascent RNA and block extension of very short transcripts, which makes them particularly effective against initiation-stage complexes. Clinically, rifamycins are central in treatment of tuberculosis and other bacterial infections, but resistance can arise through mutations in RNAP that reduce drug binding.

**Table 20.4. Bacterial RNAP Inhibitor Classes.** The table lists major bacterial RNAP inhibitor classes, their target sites, the transcription steps they block, and key selectivity and resistance considerations.

| Inhibitor Class | Target Site | Blocked Step | Selectivity Issue | Resistance Risk |
| --- | --- | --- | --- | --- |
| **Rifamycins** | RNA-exit channel near the active site (β subunit) | Extension of very short transcripts (~2–3 nt); blocks early elongation | Bacterial-specific pocket not present in human nuclear RNAPs | Single amino-acid substitutions in rpoB; widespread in M. tuberculosis clinical isolates |
| **Switch-region inhibitors** | Switch regions of β and β′ that control clamp motion | Nucleic-acid loading and clamp conformational change required for initiation | Exploits bacterial-specific switch geometry relative to eukaryotic polymerases | Mutations at targeted switch contacts; fitness cost variable |
| **Bridge-helix or trigger-loop inhibitors** | Bridge helix or trigger loop near the active site | Translocation or catalytic step of the nucleotide addition cycle | Conserved elements limit the selectivity window; bacterial specificity must be engineered | Active-site mutations confer resistance with variable fitness cost |
| **Secondary-channel inhibitors** | Secondary channel (NTP entry path into active site) | NTP access to the active site; also inhibits Gre-factor-stimulated cleavage | Channel geometry differs across bacterial species, offering some selectivity | Mutations lining the secondary channel can reduce inhibitor binding |
| **Sigma-associated inhibitors** | σ–core interaction surfaces or σ–DNA contacts | Sigma binding to core RNAP or sigma-dependent promoter recognition | σ-core interface differs from eukaryotic counterparts; some bacterial σ diversity complicates spectrum | Mutations at σ or core contact surfaces; resistance mapping ongoing |

Other inhibitor classes target different RNAP surfaces or states. Some compounds bind the switch region and interfere with clamp motion or nucleic-acid loading. Some target the bridge helix, trigger loop, or secondary channel. Some aim to disrupt sigma-core interactions or sigma-dependent initiation. The existence of multiple targetable surfaces reinforces a mechanistic point: RNAP is not a rigid enzyme with one active site. Its regulatory and catalytic cycle exposes conformations that can be selectively stabilized or blocked.

Drug selectivity is a layered problem. Bacterial RNAP differs from human nuclear RNA polymerases, which creates a therapeutic window. But a useful antibiotic must also cross bacterial envelopes, avoid efflux, retain activity against diverse bacterial RNAP variants, avoid host toxicity, and minimize rapid resistance. Gram-negative outer membranes, intracellular pathogens, biofilms, slow-growing states, and efflux systems can all reduce apparent potency. A compound that inhibits purified RNAP at low concentration is not automatically a good antibiotic.

Structural biology has transformed RNAP from a textbook cartoon into a state-resolved enzyme. X-ray crystallography and cryo-electron microscopy have captured core enzymes, holoenzymes, initiation complexes, elongation complexes, paused complexes, factor-bound complexes, antitermination complexes, and inhibitor-bound complexes. A recent cryo-EM structure of Porphyromonas gingivalis RNAP illustrates how pathogenic bacterial RNAP structures can combine conserved catalytic architecture with organism-specific surfaces relevant to drug discovery. Structures are especially powerful when paired with mutants, inhibitors, or nucleic-acid scaffolds that isolate one state.

Kinetic methods remain equally important. Transcription is a sequence of transient states, and many states are too short-lived or heterogeneous to infer from one structure. Stopped-flow fluorescence, rapid quench assays, nucleotide-walk experiments, single-molecule optical tweezers, magnetic tweezers, fluorescence resonance energy transfer, and time-resolved sequencing can measure rates, branching, dwell times, and force responses. Kinetic modeling asks which set of state transitions can explain the observations. The strongest mechanistic studies connect structural states with rate constants and cellular consequences.

Genetic methods add another layer. Mutations in RNAP subunits, sigma factors, promoters, terminators, Rho, Nus factors, Gre factors, or resistance-associated residues can reveal which surfaces and states matter in cells. Suppressor mutations are especially informative because they can connect apparently separate steps. For example, a mutation that changes pausing may alter termination, antitermination, or promoter-proximal behavior. The limitation is that genetic effects can be indirect; a mutation that changes growth rate, RNAP abundance, or stress signaling can secondarily alter transcription.

High-throughput and genomic methods are now important for bacterial transcription. Nascent RNA sequencing, native elongating transcript sequencing, transcript-end mapping, global run-on approaches, ribosome profiling, and chromatin-like occupancy assays can show where RNAP, RNA ends, ribosomes, or factors are positioned across genomes. These methods help connect purified mechanisms to cellular transcription. Their interpretation requires caution because RNA processing, RNA decay, library bias, crosslinking, extraction, and alignment rules can affect apparent RNAP density or transcript boundaries. Chapters [129](chapter1154.md), [135](chapter1123.md), and [139](chapter1126.md) discuss these measurement issues in more detail.

## Biological Contexts Across Bacteria

The basic RNAP catalytic machinery is conserved across bacteria, but transcription systems vary in sigma-factor repertoires, promoter grammars, regulatory proteins, genome organization, growth rates, and environmental pressures. Escherichia coli is a useful model, but it is not a universal proxy for all bacterial transcription. Gram-positive bacteria, cyanobacteria, actinobacteria, spirochetes, intracellular pathogens, endosymbionts, and extremophiles can differ in promoter architecture, factor usage, RNAP-associated proteins, transcription-translation coupling, and stress regulation.

Alternative sigma factors are one major source of diversity. A bacterium with many environmental transitions may encode numerous sigma factors and anti-sigma systems. A streamlined endosymbiont may have fewer regulatory options. Sporulating bacteria use sigma-factor cascades that partition transcription programs by developmental stage and compartment. Pathogens can use alternative sigma factors to control virulence, envelope stress, secretion systems, or survival in host environments. The conserved core enzyme is therefore embedded in species-specific regulatory logic.

Transcription and translation are tightly coupled in many bacteria because there is no nuclear envelope separating nascent mRNA synthesis from ribosome access. A ribosome can begin translating an mRNA while RNAP is still transcribing downstream sequence. This coupling affects Rho termination, mRNA folding, attenuation, riboswitch decisions, and transcriptional polarity. However, coupling strength varies by organism, gene, growth condition, and mRNA context. Some bacteria show more spatial or temporal separation than the simplest textbook model suggests.

DNA topology also matters. Bacterial chromosomes are supercoiled and organized by nucleoid-associated proteins. Transcription generates positive supercoils ahead of RNAP and negative supercoils behind it. Supercoiling can influence promoter opening, elongation, pausing, R-loop formation, and conflicts with replication. Topoisomerases, nucleoid proteins, and transcription factors therefore shape RNAP behavior indirectly by changing the physical substrate.

Stress physiology provides concrete examples of distributed control. During nutrient limitation, alarmone signaling and regulatory factors can alter initiation at rRNA and amino acid biosynthesis promoters. During heat shock, alternative sigma factors redirect transcription toward protein-folding and damage-response genes. During envelope stress, sigma-factor pathways activate membrane and periplasmic repair genes. In each case, transcription output changes because RNAP availability, sigma competition, promoter kinetics, and elongation or termination factors are altered together.

## Technology, Computational, Clinical, and Engineering Links

Bacterial RNAP is used in biotechnology and synthetic biology even when the enzyme itself is not the final product. Bacterial promoters, terminators, riboswitches, and sigma-factor systems are standard parts for engineered gene circuits. Understanding promoter escape, pausing, and termination helps explain why a synthetic promoter that looks strong by sequence may behave unpredictably in a plasmid, genome integration site, or different bacterial host. Terminator design must consider hairpin stability, U-tract strength, readthrough, and upstream sequence context. Riboswitch design must consider RNAP speed and ligand-binding time, not only aptamer affinity.

Computational models of bacterial transcription range from promoter prediction to kinetic simulations. Promoter-prediction models use sequence features such as -35 and -10 elements, spacer length, discriminator regions, DNA shape, and regulatory motifs. Kinetic models represent initiation and elongation as state transitions with rate constants. Genome-scale models may incorporate RNAP allocation, sigma competition, supercoiling, ribosome coupling, and termination. These models can be useful, but they fail when the input features do not capture the real limiting step. A promoter model trained on one species or growth condition may not transfer to another because sigma factors, regulators, and DNA topology differ.

Clinically, bacterial RNAP matters because transcription inhibitors are antimicrobial drugs and because resistance mutations can affect treatment outcome. RNAP-targeting compounds also help define bacterial transcription states in cells. Rifamycin resistance illustrates a general principle of RNA-pathway drugs: target engagement, resistance, permeability, bacterial physiology, and combination therapy are inseparable. [Chapter 160](chapter1143.md) discusses antiviral polymerase-pathway drugs, while [Chapter 161](chapter1144.md) discusses RNA-targeting small molecules more broadly; RNAP inhibitors are a bacterial transcription counterpart to those pharmacological themes.

## Recent Consensus

Current mechanistic consensus treats bacterial transcription as a state-based kinetic process. RNAP does not simply bind a promoter, make RNA at constant speed, and fall off at a terminator. It samples promoter complexes, opens DNA, makes abortive products, escapes or fails to escape, elongates with sequence-dependent rates, pauses, backtracks, proofreads, responds to factors, and terminates through competing pathways.

The consensus also emphasizes that promoter output is distributed across initiation steps. Sigma factors determine promoter specificity, but promoters differ in binding, opening, open-complex stability, start-site selection, abortive initiation, and escape. Regulatory proteins can act at any of these steps. Therefore, promoter strength and regulation should be described mechanistically whenever possible rather than compressed into a single on or off label.

For elongation, the consensus is that pausing is intrinsic to RNAP function and can be regulatory, corrective, or incidental depending on context. Backtracking is a conserved route to arrest and proofreading. Nascent RNA structures and elongation factors can alter the energy landscape of the elongation complex. Termination is a kinetic competition involving RNAP speed, RNA folding, Rho access, factor binding, translation, and nucleic-acid stability.

For methods, the consensus is that structural biology and kinetics must be combined. Structures reveal contacts and conformations. Kinetics reveals rates and branching. Genetics reveals cellular relevance. Genomic methods reveal positions and condition dependence. A convincing model of transcription mechanism usually needs more than one evidence class.

## Open Questions, Controversies, Deprecated Models, and Common Misconceptions

Open questions:

- At the promoter level, the field still needs better quantitative models that predict promoter output from sequence, sigma competition, NTP concentration, supercoiling, regulator occupancy, and growth condition.
- At the elongation level, many pause sites detected in vitro or by sequencing remain difficult to classify as regulatory, incidental, or method-specific.
- At the termination level, Rho loading rules and antitermination mechanisms vary across organisms and transcripts.
- At the drug-discovery level, many RNAP inhibitors fail because biochemical potency does not translate into whole-cell activity or acceptable resistance profiles.

Common misconceptions:

- "Core RNAP is the same as holoenzyme." Core RNAP contains catalytic elongation capacity, while holoenzyme contains sigma-dependent promoter-recognition capacity.
- "Promoter consensus strength equals promoter output." Promoter output is a kinetic composite of binding, DNA opening, start-site selection, abortive initiation, promoter escape, and early elongation.
- "Abortive initiation is an experimental artifact." Abortive initiation is a normal outcome of promoter-bound initial transcription for many promoters, although its frequency can vary by assay and condition.
- "Elongation is uniform once RNAP escapes." Elongation includes pauses, backtracking, factor responses, fidelity checkpoints, RNA-structure effects, and termination decisions.
- "Every mapped pause is regulatory." A pause requires cellular evidence, sufficient lifetime, perturbation response, and functional consequence before it should be called regulatory.
- "RNAP inhibitors are interchangeable transcription blockers." Different inhibitors act at different structural sites and transcription states, so their effects depend on timing, resistance mutations, permeability, and bacterial physiology.

Deprecated or weakened claims:

- Some older textbook simplifications should be treated as deprecated or incomplete. The model in which sigma always dissociates immediately after initiation is too rigid. Sigma release is context-dependent, and sigma can influence early elongation. The model in which intrinsic terminators are simple hairpin-plus-U-tract switches is incomplete because termination depends on pausing and folding kinetics. The model in which Rho only terminates abnormal transcription is incomplete because Rho also participates in normal regulation, polarity, and suppression of pervasive transcription.