1. A Central Role of the RNA Polymerase Trigger Loop in Active-Site Rearrangement during Transcriptional Pausing 
Innokenti Toulokhonov, Jinwei Zhang, Murali Palangat, Robert Landick 
Molecular Cell 
Volume 27, Issue 3, Pages (August 2007)
DOI: /j.molcel
Copyright © 2007 Elsevier Inc. Terms and Conditions

2. Figure 1 Nucleotide Addition Cycle and Transcriptional Pausing
(A) Steps in the nucleotide addition cycle. NTP (green), complexed with Mg2+II (yellow sphere), binds first to the E site prior to loading into the A subsite, which binds Mg2+I (Kd ≈ 100 μM versus >10 mM for Mg2+II) (Sosunov et al., 2003). NTP binding or movement toward the A site may weaken 3′ nt contacts and promote translocation to the product site (P) (Temiakov et al., 2005). Alternatively, (possibly slower) translocation may precede NTP binding. A rate-limiting conformational change is thought to close the active site and align the RNA 3′ OH, NTP, and Mg2+ ions for catalysis. Active-site reopening then allows PPi release. Pausing occurs when an active-site rearrangement temporarily inhibits the NAC, creating the elemental paused conformation (whether translocation is possible in elemental PTC is unresolved). At the his pause site, the short-lived elemental pause is stabilized by interactions of the pause RNA hairpin with the flap and clamp domains, spacer nucleotides, the RNA:DNA hybrid in the main channel, and the downstream DNA duplex (depicted in PTC cartoon; blue, RNAP; red, RNA; black, DNA).
(B) A structural model of the pretranslocated EC (see the Supplemental Experimental Procedures). The BH (cyan) and TL (orange) are capable of movement (arrows) based on different conformations observed in different crystal structures. The sequence insertion SI3 is specific to a lineage of γ-proteobacteria, connects to the TL by a flexible linker, and forms two domains whose orientation on RNAP is not known (Chlenov et al., 2005). βDloopII (semitransparent dark green) is shown with enlarged black Cα spheres corresponding to the 5101 substitutions P560 and T563 (partially obscured). BH and TL were modeled as in Protein Data Bank (PDB) IDs 1I6H and 1ZYR, respectively (Gnatt et al., 2001; Tuske et al., 2005).
(C) α-helical hairpin conformation of the TL observed in crystal structures with NTP bound in the A site (Vassylyev et al., 2007b; Wang et al., 2006). βDloopII is omitted for clarity. BH and TL modeled as in PDB ID 2E2H.
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2007 Elsevier Inc. Terms and Conditions

3. Figure 2
Neither the 5101 Pause-Suppressing Mutant nor 3′ thioU Alters the Pretranslocated Register of the his PTC
See the Supplemental Data for experimental details. Errors are SD of triplicate experiments.
(A) Conventional pause assay comparing wild-type and 5101 mutant RNAP pausing. E, maximal fraction of RNAP that recognizes the pause. t1/2, half-life of escape for fraction of RNAP that pauses.
(B) Delay assay showing slow formation of PTCs by 5101 RNAP.
(C) Reconstitution of his PTCs and modest effect of 3′ thioU on pausing.
(D) Detection of the pretranslocated register by Fe2+-mediated RNA cleavage.
(E) Crosslinking of 3′ thioU to wild-type and 5101 mutant RNAP reveals different contacts to the β′ and β subunits.
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2007 Elsevier Inc. Terms and Conditions

4. Figure 3 Sites of his Pause 3′ nt Crosslinking to β′ and β
(A) 3′ nt crosslinks to β′. C-terminal partial CNBr fragments are shown by lines to the sites of CNBr cleavage, with the level of 32P from the 3′ nt crosslink (normalized to 32P-end-labeled CNBr fragments generated under the same conditions; Experimental Procedures) represented as bars above the cleavage sites. Fragments lacking numbers were not resolved on the denaturing polyacrylamide gel used for mapping (Figures S2–S4). The crosslink location is indicated by the magenta sequence below the subunit map.
(B) 3′ nt crosslinks to β. The relative level of 32P from the 3′ nt crosslink in N-terminal (5101 RNAP) and C-terminal (wild-type RNAP) CNBr fragments is depicted as described for (A). The minor β1232–1243 crosslink (gray) would place the RNA 3′ nt near the RNA exit channel and probably resulted from nonspecific crosslinking of a highly reactive region of RNAP. Inexplicable crosslinking in or near this region also occurred in experiments designed to map downstream DNA contacts, even though it is more than 40 Å from the downstream DNA (Nudler et al., 1996).
(C) View of the active site with folded TL illustrating the sites of crosslinking to the TL and E site in magenta and the location of a paired or frayed 3′ nt. RNA, red. DNA, black. Mg2+I, yellow. Starburst marks the location of the photoreactive 4-thio moiety. βDloopII is omitted for clarity. RNA and DNA modeled as in ScRNAPII EC (PDB ID 1I6H) (Gnatt et al., 2001). BH and TL are modeled as in ScRNAPII EC with bound NTP (PDB ID 2E2H) (Wang et al., 2006).
(D) Same configuration as in (C), with view rotated to highlight the location of the paired 3′ nt relative to βDloopII (green) and deduced crosslink target in magenta. The BH (cyan) and TL (orange) are rendered semitransparent.
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2007 Elsevier Inc. Terms and Conditions

5. Figure 4
Effects of βDloopII and TL Alterations on 3′ nt Location and DNA Translocation Register
Errors are SD of triplicate experiments.
(A) Relative level of 3′ nt crosslinks to βDloopII or E site in β measured by trypsin cleavage of β. The schematic shows the sites of crosslinking (Figure 3) and of trypsin cleavage. The bar graph shows the relative levels of βDloopII and E site crosslinks.
(B) Upstream and downstream boundaries of ExoIII digestion of hairpin-containing (+hp) and no-hairpin (−hp) ECs. Alterations of the active-site nucleotide are indicated in the insets. mm, template base T that forces 3′ nt mismatch. +2U, two noncomplementary 3′ Us that lock the EC in a translocation register corresponding to pretranslocated his PTC. ANH2, 3′-aminoA (with corresponding template T). GTP and UTP indicate NTPs added to stimulate NTP-coupled translocation.
(C) Effects of increasing GTP concentration on translocation register. The gel panels show the downstream ExoIII boundary at 0, 0.05, 0.25, 1, and 5 mM GTP for each RNAP. The NTP-induced shift is quantified on the plot by averaging the relative levels of 32P in the bands shown in the gel panels and calculating the shift relative to the no-GTP lane.
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2007 Elsevier Inc. Terms and Conditions

6. Figure 5 Effect of TL and BH Alterations on Pause Escape Rate
(A) Pause durations as measured by pause half-lives with fold increases in half-life above each bar. Colors indicate locations of alterations (orange, TL segment deleted in ΔTL; magenta, pause 3′ nt crosslink; cyan, BH) aligned with TL sequences from E. coli (Ec), T. thermophilus (Tt), S. cerevisiae (Sc), and H. sapiens (Hs) RNAPs. The dotted region indicates the location of SI3 in E. coli RNAP. Two double substitutions are marked in red. ΔSI3 is β′Δ(943–1130) (Artsimovitch et al., 2003). ΔTL is β′Δ(931–1137)Ω(Ala)3, and ΔBH is β′Δ(789–793)Ω(Gly)3. Errors are SD of triplicate experiments.
(B and C) Locations of ΔBH and A791G relative to locations of BH unfolding (B) (Vassylyev et al., 2002) or movement (C) (Wang et al., 2006), which are shown in red aligned with the cyan BH observed in an ScRNAPII EC (Gnatt et al., 2001) (PDB ID 1I6H).
(D) Location of the ΔTL relative to the TL conformations observed in the yeast RNAPII EC (Wang et al., 2006) and bacterial RNAP bound by Stl (semitransparent) (Tuske et al., 2005). The location of the 3′ nt crosslink is magenta. The portions of the trigger helices that remain in ΔTL RNAP are gray.
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2007 Elsevier Inc. Terms and Conditions

7. Figure 6 Effect of TL in Kinetics of Transcriptional Pausing
(A) Experimental scheme for pause kinetic experiments. Numbering above sequence corresponds to promoter-containing his pause template used in (G). Numbering below refers to reconstitution scaffolds (Figure S1) used in (B)–(F).
(B) Samples from pause reactions of wild-type C28 complexes conducted on a rapid quench-flow apparatus (Experimental Procedures).
(C) Samples from conventional pause reactions of ΔTL C28 complexes.
(D) Comparison of empirical rates for behavior of wild-type and ΔTL RNAPs at 1 mM GTP.
(E) Comparison of the rates of pause escape as a function of GTP concentration for wild-type RNAP with (right scale) or without (left scale) a pause hairpin. Estimated kinetic parameters are given in the insert assuming apparent cooperativity in NTP binding (solid lines; nH = 1.2 minus hairpin, 1.7 plus hairpin; see the Supplemental Experimental Procedures) but would not differ dramatically if calculated assuming hyperbolic NTP dependence (dotted line, shown for plus hairpin only). Errors are SD from replicate measurements.
(F) Comparison of the rates of pause escape as a function of GTP concentration for ΔTL RNAP with or without a pause hairpin. Simple hyperbolas are adequate to fit the observed rates. Errors are SD from replicate measurements.
(G) Effect of Stl on nucleotide addition of his PTCs and control A63 ECs.
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2007 Elsevier Inc. Terms and Conditions

8. Figure 7 TL Model of Transcriptional Pausing
The top series of images depicts steps in the NAC with possible TL (orange) and BH (cyan) movements. From left to right, substrate NTP enters by initial interaction in the E site when the TL is unfolded (BH modeled on PDB ID 1I6H; TL modeled on PDB ID 1IW7), NTP binds in the A site with contacts to the folded TL (TL and BH modeled on PDB ID 2E2H), and NTP aligned for catalysis (BH and TL modeled on PDB ID 2O5J). The vertical series of images depicts the pathway of pausing showing a hypothetical paused TL conformation (orange). The insert depicts the interactions at a hairpin pause site that may stabilize the paused TL conformation (see text). The hairpin is depicted within the RNA exit channel as dictated by crosslinks to the flap and exit channel (Toulokhonov et al., 2001); SI3 is depicted as described in the legend to Figure 1B.
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2007 Elsevier Inc. Terms and Conditions