Structural Atlas

Escherichia coli RNA polymerase structure. Most of the structural features illustrated here were described by Zhang et al.¹  and Cramer et al.²  Views of an E. coli (Eco) RNA polymerase (RNAP) transcription elongation complex (EC).³  Proteins are shown as molecular surfaces. Some highlighted structural features are shown as backbone ribbons with transparent molecular surfaces. The nucleic acids (template-strand DNA, t-DNA; nontemplate-strand DNA, nt-DNA; RNA) are shown as cartoon worms. Lineage-specific inserts Si1, Si2, and Si3 (unique to Eco RNAP, also called βi4, βi9, and β′i6)⁴  are shown as backbone ribbons without the molecular surface. The structural features illustrated here are also detailed in Table 1. (a) Front view of the EC, looking straight into the RNAP active site cleft between the clamp and βprotrusion–βlobe. The direction of transcription is from left to right. The downstream duplex DNA (on the right) is enclosed in a channel between the clamp and β′jaw, and the βlobe. (a′) Cutaway view (with obscuring features removed for clarity) showing the boxed region in (a), affording a view of structural features surrounding the RNAP active site. The active site is marked by the Mg²⁺-ion chelated by the NADFDGD motif. The growing end of the RNA forms an RNA–DNA hybrid with the t-DNA, but upstream single-stranded RNA is directed up to the left in an RNA exit channel. The switch regions (βSw3 and β′Sw2) interact with the RNA–DNA hybrid, and β′Sw2 acts as a hinge connecting the mobile clamp domain to the rest of the RNAP. The bridge-helix (β′BH) bridges across the active site cleft. The BH and trigger-loop (TL)-helices interact to form a structural unit that is central to RNAP function, coordinating interactions with t-DNA and the incoming substrate nucleoside triphosphate (NTP). The TL is a loop connecting the TL-helices, which tends to be unstructured in unliganded RNAP, but undergoes a folding transition to interact with the correct (matched) incoming nucleotide substrate.^5,6  The F-loop modulates transcription elongation by interacting with the folded TL. The TL shown here is unfolded. In Eco RNAP, the lineage-specific insertion Si3 is inserted by flexible linkers (dashed lines) in the middle of the TL. (b) The front view (a) is rotated 45° about the vertical axis as shown, giving rise to a view roughly parallel with the axis of the upstream duplex DNA (upstream view). The direction of transcription is from left to right. The α C-terminal domains (αCTDs, modeled) are linked by long, unstructured linkers (dashed lines) to the respective α N-terminal domains (αNTDs). (b′) Cutaway view (with obscuring features removed for clarity) showing the boxed region in (b) and focusing on structural elements of the clamp. The single-stranded RNA product can be seen extruding out the RNA exit channel underneath the βflap. The β′zinc-binding domain (β′ZBD) chelates a Zn²⁺-ion (green sphere). (c) The upstream view is rotated 150° about the vertical axis as shown, giving a view into the secondary channel (SC-view) through which nucleotides can travel to the active site. The direction of transcription is right to left. The β′rim-helices (β′RH) form a rim at the entrance to the SC. (d) The upstream view is rotated 90° about the horizontal axis as shown, giving rise to a view onto the β subunit (β-view). The direction of transcription is left to right. The βlobe covers the downstream duplex DNA binding channel, while the βprotrusion covers the RNA–DNA hybrid and upstream duplex DNA. Due to the long linkers connecting Si3 with the TL, Si3 sits outside the RNAP active site cleft, interacting with the β′shelf and β′jaw. (d′). Cutaway view (with obscuring features removed for clarity) showing the boxed region in (d) but rotated by 50° about the vertical axis as shown and focusing primarily on structural features of the clamp. This view shows how the β′lid acts to disrupt the RNA–DNA hybrid, directing the single-stranded upstream RNA into the RNA exit channel towards the upper left, and the t-DNA towards the upper right, where it reanneals with the nt-DNA (grey spheres model the path of unstructured bases). The RNAP active site (NADFDGD and Mg²⁺, BH, TL, and β′ jaw) are also shown (at the bottom). (d″). Cutaway view (with obscuring features removed for clarity) showing the boxed region in (d). This view shows the relationship between the RNAP active site elements (NADFDGD and Mg²⁺, β′BH, β′TL, and β′F-loop), the RNA–DNA hybrid, and the downstream duplex DNA (on the right). (e) The upstream view is rotated 90° about the horizontal axis as shown, giving rise to a view onto the clamp. The direction of transcription is left to right. The single-stranded upstream RNA (red) can be seen extruding out the RNA exit channel between the clamp, the βflap, and the β′dock.

Eco RNAP structure progressing through the transcription cycle. Structures of Eco RNAP are shown in different stages of the transcription cycle. Proteins are shown as molecular surfaces; nucleic acids are shown as atomic spheres (color-coding is shown in the key at lower right). Each row shows a different view of the complex (top row, front view; middle row, upstream view; bottom row, β-view). Each column shows a different complex in the transcription cycle. From left to right; RNAP core, RNAP holoenzyme (core + σ⁷⁰), RNAP holoenzyme open promoter complex (RPo), and finally the transcription elongation complex (RNAP core + DNA template + RNA transcript; EC). The RPo structure shown here was adapted from PDB 6OUL.³⁸  The RNAP holoenzyme and RNAP core models shown here were derived from 6OUL by removing the DNA and σ⁷⁰, respectively. The disposition of σ⁷⁰_1.1 in the RNAP holoenzyme structure was modeled from PDB 4LK1.⁷⁰  The EC structure was adapted from PDB 6ALF.³  RNAP core: The front and upstream views illustrate the overall crab-claw-like shape of the RNAP. The top pincer in this view comprises primarily the β′ subunit, with the main part of the pincer comprising the clamp domain (see Figure 1a, b, b′, c, d′ and e). The bottom pincer comprises primarily the β subunit. The large open cleft between the pincers contains the RNAP active site (marked by a Mg²⁺-ion) and will accommodate the nucleic acids in the RPo and EC structures. RNAP holoenzyme: The σ⁷⁰ structure comprises a series of structured domains (σ⁷⁰_1.1, σ⁷⁰₂, σ⁷⁰₃, and σ⁷⁰₄) linked by flexible linkers.^71–74  Upon binding to the RNAP core, σ⁷⁰_1.1 situates in the RNAP active site cleft, possibly serving to protect the RNAP active site from engaging with non-specific nucleic acids.⁷⁰  The σ⁷⁰_1.1 is ejected from the RNAP active site cleft upon the entry of promoter DNA.⁷⁵  The main functional role of σ⁷⁰₂, which binds to the β′clamp-helices (β′CH; see Figure 1b′, d′), is to initiate transcription bubble formation and recognize the promoter −10 element.⁷⁶  The σ⁷⁰₃ recognizes the promoter extended −10 element, when present.^59,77  The main functional role of σ⁷⁰₄, which comprises a canonical helix–turn–helix DNA binding domain, is to recognize the promoter −35 element.⁷²  The σ⁷⁰₄ is perched on the RNAP βflap-tip helix (see Figure 1a, b, b′, e). Binding of σ⁷⁰₂ to the β′CHs and σ⁷⁰₄ to the βflap positions the two σ⁷⁰ domains appropriately to engage the promoter −10 and −35 elements, that are typically separated by a 17 base pair spacer.^22,78,79  The σ⁷⁰₃ and σ⁷⁰₄ are linked by a long linker (the σ⁷⁰_finger) that snakes into the RNAP active site cleft near the RNAP active site, then back out through the RNA exit channel.²²  The σ⁷⁰_finger facilitates de novo initiation (initiation of RNA chain synthesis that does not require an oligonucleotide primer) by helping to properly position the t-strand DNA,^72,80,81  but upon RNA chain initiation and subsequent elongation, the σ⁷⁰_finger becomes an obstacle to RNA chain elongation that must be displaced.^6,22,82,83  RPo: Location of a duplex promoter sequence by the RNAP holoenzyme triggers a series of conformational changes as the enzyme unwinds 12 to 14 base pairs of DNA to generate the transcription bubble and loads the t-strand DNA into the RNAP active site.^30,84–87  This multi-step process culminates in the formation of the transcription competent RPo. In RPo, σ⁷⁰₄ recognizes the duplex −35 element and σ⁷⁰₂ engages the −10 element and stabilizes the upstream edge of the transcription bubble.^59,72,76  The single-stranded transcription start site is loaded into the RNAP active site, and the duplex DNA downstream of the start site is bound in the downstream duplex channel, displacing σ⁷⁰_1.1.⁷⁵  At many promoters, the αCTDs interact with AT-rich DNA upstream of the −35 element (UP-elements).^12,14–16  Additional promoter–RNAP holoenzyme interactions are detailed in Figure 3. EC: The formation of RPo allows NTP substrates to enter the RNAP active site (through the SC-channel; Figure 1c) and RNA chain extension to begin. The initial elongation of the RNA is characterized by fits and starts, marked by scrunching and abortive initiation (release of short RNA chains without release of RNAP holoenzyme from the promoter),^88–90  but ultimately the RNA product is extruded through the RNA exit channel and σ⁷⁰ is released, giving rise to the EC.

Structural determinants of RNAP holoenzyme–promoter DNA interactions. The center of the image shows the β-view of RPo (the same as Figure 2). The circled features are highlighted around the outside of the image. (a) UP-element: At some promoters, the αCTDs bind to an AT-rich sequence just upstream of the promoter −35 element.^12,14,15  The proximal αCTD (the one closest to RNAP) can make favorable interactions with σ⁷⁰₄¹⁶  as shown here. (b) −35 element (consensus ⁻³⁵TTGACA⁻³⁰; where the superscripts indicate the usual position of the element with respect to the transcription start site, which can vary):⁷⁸  In Eco, about 43% of σ⁷⁰ promoters contain an identifiable −35 element,⁹¹  making it the most prevalent promoter element after the −10 element. The −35 element sequence is recognized by σ⁷⁰₄, a canonical helix–turn-helix DNA binding domain.^72,79  (c) Extended −10 element (⁻¹⁵TG⁻¹⁴): In Eco, about 18% of σ⁷⁰ promoters contain an extended −10 element; a TG sequence spaced one base pair upstream of the −10 element.^91,92  The presence of an extended −10 element can substitute for the absence of a −35 element.⁹³  The extended −10 element is recognized by residues of σ⁷⁰₃.^59,77  Upstream of the extended −10, residues from the β′zipper interact with the DNA phosphate backbone at the −17 position of the nt-strand.^58,59  (d) −10 element (⁻¹²TATAAT⁻⁷; A₋₁₁): The −10 element is the most conserved promoter element, and the A at the second position (normally A₋₁₁) is the most conserved base of the −10 element.⁷⁸  The A_-11 is thought to be the first base that flips out of the duplex DNA stack to nucleate transcription bubble formation.^76,94–96  The flipped-out A₋₁₁ is captured in a cognate pocket of σ⁷⁰_2.⁷⁶  (e) −10 element (⁻¹²TATAAT⁻⁷; T₋₇): The last T of the −10 element (normally T₋₇) is nearly as conserved as the A₋₁₁.⁷⁸  T₋₇ also flips out of the DNA base stack and is captured in a separate pocket of σ⁷⁰_2.⁷⁶  (f) Discriminator (⁻⁶GG⁻⁵): The ‘discriminator’ is defined as the sequence between the −10 element and the start site (to −6 to −1). Characteristics of the discriminator have a profound effect on RPo stability.^97–99  The two positions just downstream of the −10 element are particularly important, where a ⁻⁶GG⁻⁵ sequence can form specific interactions with σ⁷⁰₂.⁸⁰  (g) Core recognition element (CRE: G₊₂): A G at the +2 position of the nucleotide strand can bind in a G-specific pocket of the β-subunit.⁸⁰