Structural basis of transcription inhibition by the DNA mimic protein Ocr of bacteriophage T7

Fuzhou Ye; Ioly Kotta-Loizou; Milija Jovanovic; Xiaojiao Liu; David TF Dryden; Martin Buck; Xiaodong Zhang

doi:10.7554/eLife.52125

. 2020 Feb 10;9:e52125. doi: 10.7554/eLife.52125

Structural basis of transcription inhibition by the DNA mimic protein Ocr of bacteriophage T7

Fuzhou Ye ¹, Ioly Kotta-Loizou ², Milija Jovanovic ², Xiaojiao Liu ^1,³, David TF Dryden ⁴, Martin Buck ², Xiaodong Zhang ^1,^✉

Editors: Seth Darst⁵, Cynthia Wolberger⁶

PMCID: PMC7064336 PMID: 32039758

Abstract

Bacteriophage T7 infects Escherichia coli and evades the host restriction/modification system. The Ocr protein of T7 was shown to exist as a dimer mimicking DNA and to bind to host restriction enzymes, thus preventing the degradation of the viral genome by the host. Here we report that Ocr can also inhibit host transcription by directly binding to bacterial RNA polymerase (RNAP) and competing with the recruitment of RNAP by sigma factors. Using cryo electron microscopy, we determined the structures of Ocr bound to RNAP. The structures show that an Ocr dimer binds to RNAP in the cleft, where key regions of sigma bind and where DNA resides during transcription synthesis, thus providing a structural basis for the transcription inhibition. Our results reveal the versatility of Ocr in interfering with host systems and suggest possible strategies that could be exploited in adopting DNA mimicry as a basis for forming novel antibiotics.

Research organism: E. coli

eLife digest

Bacteria and viruses have long been fighting amongst themselves. Bacteriophages are a type of virus that invade bacteria; their name literally means ‘bacteria eater’. The bacteriophage T7, for example, infects the common bacteria known as Escherichia coli. Once inside, the virus hijacks the bacterium’s cellular machinery, using it to replicate its own genetic material and make more copies of the virus so it can spread. At the same time, the bacteria have found ways to try and defend themselves, which in turn has led some bacteriophages to develop countermeasures to overcome those defences.

Many bacteria, for example, have restriction enzymes which recognise certain sections of the bacteriophage DNA and cut it into fragments. However, the T7 bacteriophage has one well-known protein called Ocr which inhibits restriction enzymes. Ocr does this by mimicking DNA, which led Ye et al. to wonder if it could also interrupt other vital processes in a bacterial cell that involve DNA.

Transcription is the first step in a coordinated process that turns the genetic information stored in a cell’s DNA into useful proteins. An enzyme called RNA polymerase decodes the DNA sequence into a go-between molecule called messenger RNA, and it was here that Ye et al. thought Ocr might jump in to interfere.

To begin, Ye et al. examined the structure of Ocr when it binds to RNA polymerase using an imaging technique called cryo-electron microscopy. Ocr has been well-studied before, its structure previously described, but not when attached to RNA polymerase. The analysis showed that Ocr gets in the way of specific molecules, called sigma factors, that show RNA polymerase where to start transcription. Ocr binds to RNA polymerase in exactly the same pocket as part of sigma factors do, which is also the place where DNA must be to be decoded to make messenger RNA. Ye et al. then performed experiments to show Ocr interfering with binding to RNA polymerase did indeed disrupt transcription. This means Ocr is quite versatile as it interferes with the RNA polymerase of the bacterial host and its restriction enzymes.

Ocr’s strategy of mimicking DNA to interrupt transcription could be adopted as an approach to develop new antibiotics to stop bacterial infections. DNA transcription is an essential cellular process – without it, no cell can replicate and survive – and RNA polymerase is already a validated target for drugs. Following Ocr’s lead could provide a new way to stop infections, if the right drug can be designed to fit.

Introduction

Bacteriophage T7 infects Escherichia coli and hijacks the host cellular machinery to replicate its genome (Studier, 1972; Krüger and Schroeder, 1981; Hausmann and Messerschmid, 1988). The T7 genome encodes 56 proteins with many functioning as structural proteins for the bacteriophage. A number of T7 proteins are known to specifically inhibit the bacterial cellular machinery. For example, proteins gp0.7, gp2 and gp5.7 inhibit cellular transcription (Cámara et al., 2010; Tabib-Salazar et al., 2018) whereas gp0.3 inhibits restriction/modification (RM) enzymes (Studier, 1975).

Gp0.3 is the first T7 gene expressed after infection and T7 variants lacking gene 0.3 were shown to have genomes susceptible to E. coli RM systems (Studier, 1975). Subsequently the 117 amino acid protein gp0.3 was named Overcome Classical Restriction (Ocr) (Krüger and Schroeder, 1981). Ocr is abundantly expressed and forms a dimer that mimics the structure of a slightly bent 20 base pair B-form DNA (Issinger and Hausmann, 1972; Walkinshaw et al., 2002) and blocks the DNA binding grooves of the type I RM enzyme, preventing the degradation and modification of the T7 genome by the host. Intriguingly, Type I RM enzymes are present in very low numbers (estimated at ~60 molecules per cell [Kelleher and Raleigh, 1994]). Since Ocr is a DNA mimicry protein, it is possible that the abundantly expressed Ocr (estimated to be several hundreds of molecules per cell at least) (Hausmann and Messerschmid, 1988) also interferes with other DNA processing systems of the host. Indeed early evidence of an interaction between Ocr and the host RNA polymerase (RNAP) was obtained using pull-down affinity chromatography (Ratner, 1974).

RNA polymerase is the central enzyme for transcription, which is a highly controlled process and can be regulated at numerous distinct functional stages (Kornberg, 1998; Decker and Hinton, 2013). The large majority of transcription regulation, however, is executed at the recruitment and initiation stage (Browning and Busby, 2004; Hahn and Young, 2011; Browning and Busby, 2016). To ensure transcription specificity, bacterial RNAP relies on sigma (σ) factors to recognise gene-specific promoter regions. E. coli has seven sigma factors which can be grouped into two classes, the σ⁷⁰ class represented by σ⁷⁰, responsible for transcribing housekeeping genes, and the σ⁵⁴ class, responsible for transcribing stress-induced genes including phage infection (Feklístov et al., 2014; Browning and Busby, 2016). Much work has yielded a detailed mechanistic understanding of how transcription directed by σ⁷⁰ and σ⁵⁴ is initiated (Zhang et al., 2012; Glyde et al., 2018). Specifically, the two large RNAP subunits β and β’ form a crab claw structure that encloses the DNA binding cleft, accommodating the transcription bubble and the downstream double-stranded (ds) DNA (Bae et al., 2015; Zuo and Steitz, 2015).

Inhibiting host transcription is widely exploited by bacteriophages including T7 on E. coli and P23-45 on T. thermophilus (Tagami et al., 2014; Tabib-Salazar et al., 2017; Ooi et al., 2018). Gp0.7, gp2 and gp5.7 of T7 were shown to inhibit bacterial RNAP (Cámara et al., 2010; Tabib-Salazar et al., 2017). Gp0.7 is a protein kinase that phosphorylates the E. coli RNAP-σ⁷⁰, resulting in transcription termination at sites located between the early and middle genes on the T7 genome, gp2 specifically inhibits σ⁷⁰-dependent transcription initiation and gp5.7 inhibits σ^S, responsible for stationary phase adaptation (Bae et al., 2013; Tabib-Salazar et al., 2018). Two proteins gp39 and gp76 of bacteriophage P23-45 were shown to inhibit T. thermophilus transcription (Tagami et al., 2014; Ooi et al., 2018). Importantly, RNAP is a validated antibacterial target and novel inhibitors of RNAP hold promise for potential new antibiotic development against antimicrobial resistance (Ho et al., 2009; Srivastava et al., 2011; Ma et al., 2016).

In this work, we investigate the potential effects of Ocr on the host transcriptional machinery. Our results with purified components show that Ocr can bind to RNAP and inhibit both σ⁷⁰ and σ⁵⁴ dependent transcription. Specifically we establish that an Ocr competes with sigma binding, thus affecting recruitment and can inhibit transcription both in vitro and in E. coli cells when over-expressed. Using structural biology, we show that the Ocr dimer directly binds to RNAP at the RNAP cleft, where sigma factors and the transcription bubble bind, suggesting that Ocr interferes with transcription recruitment by competing with sigma factor binding, and could also interfere with transcription initiation. Our work thus reveals the detailed molecular mechanisms of how Ocr could potentially affect the host through inhibiting transcription in addition to its role in knocking out host restriction and modification. These new structural details of RNAP-Ocr complexes allow comparisons with known transcription initiation complexes and other phage proteins suggesting new avenues to be exploited for inhibiting bacterial transcription and combating antibiotic resistance.

Results

Ocr interacts with RNAP with high affinity

An early study by Ratner (1974) showed that upon T7 infection, Ocr protein was detected in pull-down experiments using host RNAP as bait, suggesting that Ocr could associate with the host transcriptional machinery. To assess if Ocr can engage with RNAP directly, we carried out in vitro interaction studies using purified components (Figure 1, Figure 1—figure supplement 1). We observed that purified Ocr co-eluted with RNAP in gel filtration experiments, suggesting a tight interaction of RNAP and Ocr sufficient to withstand the lengthy elution time of the complex during chromatography (Figure 1A–C). The complex persisted for the order of minutes during chromatography, suggesting a stable complex. To quantify the interactions, we carried out binding experiments between RNAP and Ocr using microscale thermophoresis (MST). Our results show the binding affinity (dissociation constant, Kd) is in the order of 10 nM, similar to that of RNAP with sigma factors (Figure 1D–F). These results thus confirm the basis for the RNAP-Ocr interaction originally observed (Ratner, 1974). Our observed affinity of Ocr for RNAP is about three orders of magnitude lower than the values reported for the association of Ocr with the EcoKI type I restriction modification enzyme (Atanasiu et al., 2002) as would be expected if the type I RM enzyme was the primary target for Ocr.

Figure 1. — (A) Gel filtration chromatography profiles of RNAP (green), Ocr (blue) and RNAP-Ocr complex (orange). Two peaks correspond to RNAP-Ocr complexes (B) SDS-PAGE of the corresponding fractions from (A) as indicated by coloured dashed lines in (A) and the purified RNAP and Ocr as comparison, the same below in (C). (C) Western blotting of samples from RNAP-Ocr gel filtration fractions to verify the presence of Ocr co-eluting with RNAP, here only Ocr is his-tagged. Two fractions under each peak (as coloured dash arrows shows) in gel-filtration is loaded. (**D–F**). MST experiments measure the dissociation constants (Kd) of Ocr, σ⁷⁰ and σ⁵⁴ with RNAP. These Kd values represent maximum values.

Figure 1—figure supplement 1. — (A) Gel filtration chromatography profiles of RNAP (green), Ocr (blue) and RNAP-Ocr complex (orange). Two peaks correspond to RNAP-Ocr complexes (B) SDS-PAGE of the corresponding fractions from (A) as indicated by coloured dashed lines in (A) and the purified RNAP and Ocr as comparison, the same below in (C). (C) Western blotting of samples from RNAP-Ocr gel filtration fractions to verify the presence of Ocr co-eluting with RNAP, here only Ocr is his-tagged. Two fractions under each peak (as coloured dash arrows shows) in gel-filtration is loaded. (**D–F**). MST experiments measure the dissociation constants (Kd) of Ocr, σ⁷⁰ and σ⁵⁴ with RNAP. These Kd values represent maximum values.

Structures of Ocr in complex with RNAP

In order to understand how Ocr interacts with RNAP, we subjected the purified complex from size-exclusion chromatography to cryo electron microscopy (cryoEM) single particle analysis. After several rounds of 2D classification, which allowed us to remove ice contaminations, very large particles or particles much smaller than RNAP, the remaining particles were subject to 3D classifications using the structure of the RNAP filtered to 60 Å as an initial model (PDB code 6GH5, Figure 2—figure supplement 1). Two classes with clear density for Ocr were refined to a global resolution 3.7 and 3.8 Å respectively although the quality is not uniform in the reconstruction, as demonstrated in the resolution distribution. Two resolution zones seem to be evident, around 3–4 Å and around 7–8 Å, reflecting the structural features of the complex (some domains are mobile) and partly due to preferential orientations of the particles (Figure 2, Figure 2—figure supplement 2 and Figure 2—figure supplement 3). The electron density for RNAP is clearly resolved and two distinct structural models of RNAP (taken from PDB code 6GH6 and 6GH5) can be readily fitted into the two reconstructions (Figure 2 and Figure 2—figure supplement 3A). The remaining density region that is not accounted for by RNAP can accommodate an Ocr dimer (Figure 2—figure supplement 3A). The density for Ocr is of sufficient quality that allows accurate positioning of the Ocr atomic structure (PDB code 1S7Z) (Figure 2—figure supplement 3B–3C). The regions with the highest resolution within the reconstruction are in the RNAP core (Figure 2—figure supplement 2), where density for many side chains is clearly visible (Figure 2—figure supplement 3C).

Figure 2. — (A) ‘wide-clamp’ mode. (B) ‘narrow-clamp’ mode. RNAP and σ⁷⁰ were shown as surface, Ocr shows as cartoon. The colour key is below the figure. α-grey, β-pale cyan, β’-wheat, ω-slate, σ⁷⁰-palegreen, Ocr–magenta (proximal subunit) and yellow (distal subunit), non-template strand DNA(NT)-sky blue, template strand DNA (T)-black. (C) RNAP-σ⁷⁰ open promoter complex (PDB code 6CA0).

Figure 2—figure supplement 1. — (A) ‘wide-clamp’ mode. (B) ‘narrow-clamp’ mode. RNAP and σ⁷⁰ were shown as surface, Ocr shows as cartoon. The colour key is below the figure. α-grey, β-pale cyan, β’-wheat, ω-slate, σ⁷⁰-palegreen, Ocr–magenta (proximal subunit) and yellow (distal subunit), non-template strand DNA(NT)-sky blue, template strand DNA (T)-black. (C) RNAP-σ⁷⁰ open promoter complex (PDB code 6CA0).

During 3D classification, it is clear that a number of different conformational and structural classes exist within the dataset and two distinct classes were identified due to their widely differing RNAP clamp conformations (Figure 2—figure supplement 1). Bacterial RNAP consists of five subunits forming a crab claw shape with the two large subunits β and β’ forming the claws (or clamp) which enclose the RNAP channel that accommodates the transcription bubble and downstream double-stranded (ds) DNA. The RNAP clamp is highly dynamic and single-molecule Forster Resonant Energy Transfer data have shown that the RNAP clamp adapts a range of conformations from closed to open involving more than 20 Å movement and 20^o rotation of the clamp (Chakraborty et al., 2012). Our recent structural work showed that specific clamp conformations are associated with distinct RNAP functional states with the widely opened clamp conformation associated with the DNA loading intermediate state (Glyde et al., 2018).

In both of the distinct RNAP-Ocr complex structures, Ocr remains as a dimer upon binding to RNAP (Figure 2A–B) and the Ocr dimers are slightly less bent compared to that of an Ocr alone crystal structure (Figure 2—figure supplement 4). In one structure, the entire Ocr dimer inserts deeply into the RNAP cleft, occupying the full RNAP nucleic acid binding channel including where the transcription bubble and the downstream DNA resides in the transcriptional open and transcribing complex (Figure 2A, Figure 2C, Figure 3A). In this interaction mode, the RNAP clamp is wide open and both Ocr subunits are involved in the interactions (Figure 2A, Figure 3B, Figure 2—figure supplement 3), we thus denote it as the ‘wide clamp structure’. The second structure involves only one of the two Ocr subunits within the dimer contacting RNAP, occupying only the downstream DNA channel. Here the RNAP clamp is less open and we denote this structure the ‘narrow clamp structure’ (Figure 2B, Figure 2—figure supplement 3).

Figure 3. — (**A–B**) Comparison with initial transcribing complex structure (PDB 4YLN). (C) Ocr resides in the positively charged RNAP channel. (**D–E**) Detailed charge distributions of RNAP and Ocr in the interacting regions. Blue – positive charge, red – negative charge.

In the wide clamp structure, the RNAP clamp conformation is very similar to the conformation observed in the transcription initiation intermediate complex where the clamp is wide open and DNA is partially loaded into the cleft Glyde et al., 2018). In this structure, the Ocr dimer inserts into the RNAP channel and each of the two Ocr subunits, which we term proximal and distal, and makes distinct interactions with RNAP. The proximal subunit is deeply embedded in the RNAP cleft and occupies the space for the template strand DNA, the complementary newly synthesised RNA and the non-template strand DNA (Figure 2 and Figure 3, magenta [Bae et al., 2015; Zuo and Steitz, 2015; Glyde et al., 2018]). The distal Ocr subunit occupies the position of the downstream DNA (Figure 2 and Figure 3A, yellow). The negatively charged surface of the proximal Ocr complements the highly positively charged surface of the surrounding RNAP and fits snugly into the channel formed by β and β’ subunits (Figure 3C–E). The distal Ocr subunit mimics the downstream DNA in its interactions with the β’ clamp. Superposition of the β’ clamp in the RNAP-Ocr structure with that of initial transcribing complex shows the distal Ocr monomer aligns with downstream dsDNA in the initial transcribing complex (Figure 3B). However, the Ocr dimer has a rigid conformation that resembles a bent B-DNA. Consequently, in order to maintain the interactions between both Ocr subunits and RNAP, the β’ clamp has to open up (to ~58 Å in RNAP-Ocr from ~35 Å in RPo) to accommodate the Ocr dimer compared to RPo that accommodates a transcriptional bubble and dsDNA (Figure 2).

In the narrow clamp mode, the proximal Ocr subunit interacts with the β’ clamp, again mimicking downstream DNA in RPo although the clamp is more open (~45 Å) compared to that in RPo (~35 Å) (Figure 2B–C). Interestingly, although an Ocr monomer surface mimics that of a dsDNA in the overall dimension and the negative charge distributions, the Ocr subunit in the narrow clamp structure does not exactly overlay with the downstream dsDNA in RPo when RNAP is aligned on its bridge helix, the highly conserved structural feature that is close to the active centre and connects β and β’ clamps (Figure 4A). Instead Ocr is shifted upwards compared to the dsDNA in RPo (Figure 4A). Inspecting the structures of Ocr and RNAP explain the differences. The downstream binding channel of RNAP consists of β and β’ clamps on two sides and the β’ jaw domain as a base, providing a positively charged environment on three sides, required to accommodate and engage with the negatively charged DNA during transcription (Figure 4B). Although an Ocr monomer has a largely negatively charged surface, there are also positively charged patches on the Ocr surface, especially those facing the β’ jaw domain (Figure 4C). In order to maintain the interactions between Ocr and the β’ clamp and to overcome charge repulsions with the β’ jaw, the β’ clamp is opened up, shifting Ocr upwards and away from the jaw domain.

Figure 4. — (A) Two orthogonal views of RNAP-Ocr narrow-clamp structure (RNAP shows as surface and Ocr shows as cartoon (magenta and yellow), overlaid with σ⁷⁰ open complex structure (PDB code 6CA0). (B) Surface charge distribution of RNAP showing positively charged surfaces in the cleft where Ocr binds. Magenta ellipse – where Ocr binds. (C) Surface charge distribution of Ocr facing the bottom of the cleft including β’ jaw domain. Magenta ellipse – area of Ocr that is inside the cleft showing both negative and positively charged areas. Blue – positively charged, red – negatively charged.

Ocr inhibits transcription in vitro and in cells

In order to understand the significance of Ocr binding to RNAP, we investigated the effects of Ocr on in vitro transcription using short-primed RNA (spRNA) assays performed on various promoter DNAs (see Materials and methods). To determine the effect of Ocr on RNAP-σ⁷⁰ transcription activity we performed in vitro transcription assays using different promoters and promoter variants such as linear σ⁷⁰ promoter, or lacUV5 promoter (Cámara et al., 2010) as well as a T7A1 supercoiled promoter (Figure 5A–B, Figure 5—figure supplement 1). Our data reveal that adding 5 μM Ocr decreased activity of RNAP-σ⁷⁰ on the different promoters in the range of 70–10%. On the linear promoter in the presence of 5 μM Ocr the amount of spRNA was significantly reduced (to 10–30%) when Ocr was added prior to or after the holoenzyme formation but before DNA was added (Figure 5A-B I-II). However, the effect of Ocr after DNA was added was less pronounced (Figure 5A–B III). Ocr did not abolish transcription on the supercoiled T7A1 promoter at the same extent as on the linear promoters, indicating a role for DNA topology and perhaps the energy barriers of DNA opening in open complex formation and the promoter sensitivity to Ocr (Figure 5—figure supplement 1C I-III).

Figure 5. — (**A-B**) spRNA experiments on σ⁷⁰ using two different promoter DNAs. Reaction schematics are shown with I, II, III representing experiments when Ocr is added during different stages of transcription initiation reaction. A control lane without Ocr is also shown. (C-D) using a reporter assay, we show that over-express of Ocr in *E. coli* cells inhibits cell growth (C) and β-galactosidase activities (D). At least three independent experiments were performed, each with 3–5 technical replicates.

Figure 5—figure supplement 1. — (**A-B**) spRNA experiments on σ⁷⁰ using two different promoter DNAs. Reaction schematics are shown with I, II, III representing experiments when Ocr is added during different stages of transcription initiation reaction. A control lane without Ocr is also shown. (C-D) using a reporter assay, we show that over-express of Ocr in *E. coli* cells inhibits cell growth (C) and β-galactosidase activities (D). At least three independent experiments were performed, each with 3–5 technical replicates.

The effect of Ocr on the major variant RNAP-σ⁵⁴ dependent transcription was also assessed using different variants of Sinorhizobium meliloti nifH promoter including linear and supercoiled DNA (Chaney and Buck, 1999; Glyde et al., 2018), (Figure 5—figure supplement 1A,B). The phage shock protein F (PspF_1-275) was used as an activator of RNAP-σ⁵⁴ transcription to promote open complex formation (see Materials and methods). Again, on both linear and supercoiled nifH promoters the RNAP-σ⁵⁴ complexes were sensitive to Ocr (Figure 5—figure supplement 1).

Our results show that Ocr can efficiently inhibit transcription in vitro. We next tested to see if expression of Ocr in cells could indeed exhibit an effect in cells. We used E. coli MG1655 cells expressing Ocr from a plasmid vector under the control of an arabinose-inducible promoter and a range of growth conditions (Supplementary file 1).

An Ocr-mediated growth inhibition effect was indeed observed in minimal modified M9 media (Figure 5C). The effects were further assessed using the chromosomal β-galactosidase gene as a reporter. During the 5 hr period following induction with IPTG, a ~ 10 fold decrease of normalised β-galactosidase activity was observed in E. coli cells expressing Ocr compared to cells that do not express Ocr (Figure 5D). Moreover, at 15 hr post-induction, upon induction, there was a 4-fold increase of normalised β-galactosidase activity in cells not expressing Ocr, while the β-galactosidase activity in cells expressing Ocr was lower overall and the observed increase upon induction was less than 2-fold (Figure 5—figure supplement 2A). Subsequent measurements of lacZ mRNA using reverse transcription followed by quantitative polymerase chain reaction (RT-qPCR) showed that the observed β-galactosidase activity is concomitant with lacZ mRNA levels (Figure 5—figure supplement 2B), confirming that Ocr was inhibitory for transcription in cells.

Ocr inhibits RNAP recruitment and open complex formation but does not disrupt pre-formed open complexes

Since Ocr interfered with transcription, we wanted to investigate whether it does so through interfering with RNAP recruitment by sigma and/or other processes required for transcription. Specifically, we probed the ability of Ocr in interfering with RNAP- σ holoenzyme and holoenzyme-promoter complex formation (recruitment) and pre-formed open complex, which is competent for transcription, thus allowing us to assess the effects on elongation.

Using purified components and gel mobility shift assays, Ocr was observed to inhibit RNAP-σ⁷⁰ holoenzyme formation in a concentration dependent fashion and irrespective of whether Ocr is added before or after σ (Figure 6A and B, comparing lane 6 with 7–10). Direct competition experiments with pre-formed RNAP-σ⁷⁰ holoenzyme show that appearance of a RNAP-Ocr complex accompanies the disappearance of the RNAP-σ⁷⁰ holoenzyme, (Figure 6A and B, comparing lane 6 with lanes 11–12). Indeed, MST experiment of Ocr binding to RNAP-σ⁷⁰ revealed that Ocr can displace σ⁷⁰ with an apparent Kd of ~20 nM (Figure 6—figure supplement 1C). This is also consistent with data showing that RNAP has similar affinities for Ocr and σ (Figure 1D–F).

Figure 6. — (A and B) Native PAGE shows Ocr can inhibit and disrupt RNAP-σ⁷⁰ holoenzyme formation. Coomassie blue staining (A) and Western blots against His-tagged σ⁷⁰ (B). (A and B) reaction schemes are indicated above. E represents RNAP core enzyme; I, II, III indicate the point when Ocr was added during the reactions. In Reaction III, holoenzyme was preformed and purified by size exclusion chromatography first. Protein concentrations are shown in the reaction schemes. In lanes 7, 9, 11, a 1:1 molar ratio of Ocr to σ was used, whereas in lane 8, 10, 12, a 4:1 molar ratio of Ocr to σ was used. The exact Ocr concentrations are labelled above each lane. (C and D) σ⁷⁰ open complex formation with different promoter DNA as assayed by native-PAGE gels and stained for DNA. Reaction schemes are indicated above. I, II, III indicate the point when Ocr was added during the reactions. The same experimental conditions were used as in vitro transcription assays in Figure 5, with either 0.4 µM or 5 µM final concentration Ocr added for each reaction, as indicated in each lane. Reaction in the absence of Ocr (w/o Ocr) was used as a control. All reactions include 100 µg/ml heparin to ensure only open complexes are captured.

We next assessed the ability of Ocr in inhibiting recruitment to the promoter DNA by monitoring the amount of DNA bound holoenzymes using two different promoters (Figure 6C–D). Ocr clearly inhibits the formation of a RNAP-σ⁷⁰-DNA complex in a concentration dependent manner. Adding 5 μM Ocr either before or after σ addition resulted in ~30% promoter complex formation (Figure 6C–D,I–II). Adding Ocr after DNA has only limited effects (Figure 6C–D III). This is consistent with the higher binding affinity between RNAP-σ and DNA (0.1 nM for σ⁷⁰ and 3 nM σ⁵⁴) (Figure 6—figure supplement 1) compared to RNAP-Ocr (Figure 1D). These results are in excellent agreement with those in vitro transcription assay (Figure 5).

Similar effects were observed for RNAP-σ⁵⁴ (Figure 6—figure supplement 1). Due to the lower sensitivity of the diffused σ⁵⁴ band to Coomassie blue staining, the displaced free σ⁵⁴ from the RNAP-σ⁵⁴ holoenzyme was observed by western blotting with an antibody to the His-tag on σ⁵⁴.

Together our binding experiments and in vitro transcription assays support the proposal that Ocr inhibits RNAP recruitment to the promoter DNA by directly competing with σ for RNAP binding. However, Ocr has only limited effects on pre-formed RNAP-σ-DNA complexes. This is similar to the interactions of Ocr with a type I RM enzyme where Ocr only affects the enzyme when not bound to DNA (Bandyopadhyay et al., 1985).

Discussion

Ocr is a bifunctional protein

Our data here characterise a new activity for Ocr in inhibiting the host transcriptional machinery in addition to its established role in inhibiting the type I RM systems of the host. Thus, Ocr is a bifunctional DNA mimic protein.

The in vitro transcription assay and Ocr competition experiments clearly show that Ocr can inhibit σ binding and recruitment of RNAP to promoter DNA. Our data using a reporter assay also demonstrate the ability of Ocr in inhibiting in vivo the E. coli transcription machinery. The structures of Ocr-RNAP show that Ocr occupies the space where σ binds (Figure 7A–B) and where the transcription bubble and downstream DNA reside in the open complex and elongation complex, thus explaining how Ocr inhibits RNAP recruitment by σ (Vassylyev et al., 2007; Zuo and Steitz, 2015).

Figure 7. — (A) Ocr overlay with RNAP-σ⁷⁰ holoenzyme B) Complex structure of gp2 of T7 with σ⁷⁰ holoenzyme (PDB 4LLG), (C) Complex structure of gp76-gp39 of P23-45 with σ⁷⁰ holoenzyme (PDB 5XJ0), (D) Enlarged view of gp76 and σ⁷⁰. RNAP shows as cartoon and coloured as grey, σ⁷⁰ shows as surface and coloured pale green. Gp2, gp39, gp76 are show as surface and coloured as firebrick, light pink and firebrick, respectively.

Our data suggest an answer to the puzzle of why Ocr is so abundantly overexpressed immediately upon a T7 infection (Issinger and Hausmann, 1972; Studier, 1975), when its high affinity classical target, the Type I RM enzyme, is present in very low numbers (estimated at ~60 molecules per cell, [Kelleher and Raleigh, 1994]). The very tight binding between Ocr and Type I RM enzymes ensures complete inhibition of restriction and modification. It would appear from our data that the excess Ocr molecules could then start the process of shutting down host transcription by binding to any non-transcribing RNAP. An earlier study by McAllister and Barrett used host strains lacking the host restriction systems, thus eliminating the effects of Ocr on host restriction enzymes. They showed that expression of gp0.3 (Ocr) resulted in a delayed inhibition of host β-galactosidase synthesis and delayed inhibition of total RNA synthesis (McAllister and Barrett, 1977). This inhibitory effect is consistent with our studies here showing that Ocr can inhibit host transcription. Since Ocr is expressed by host RNAP, it thus takes time for a sufficient amount of Ocr to be synthesized by the host, resulting in the delayed response. Indeed, Ocr could bind to free RNAP and/or potentially compete with host σ⁷⁰ which are in the range of hundreds of molecules per cell (https://pax-db.org/dataset/511145/137/), since their binding affinities to RNAP are similar (Figure 1D–F) and Ocr can compete with σ⁷⁰ for RNAP binding and promoter DNA binding (Figure 6—figure supplement 1C). Interestingly, when Ocr and gp1.1 (a small protein with unknown function) were both expressed, the host protein synthesis was shutdown effectively, suggesting that gp1.1 and Ocr might have complementary/synergetic effects.

The bifunctionality observed in Ocr may also be shared by other DNA mimic proteins. For example, the phage T4 Arn protein targeting the McrBC restriction enzyme also has weak interaction with histone-like protein H-NS (Ho et al., 2014). Arn is very similar in shape to Ocr but since T4 requires the host RNAP throughout its life cycle it is unlikely to target RNAP as strongly as Ocr. Additional functions may also exist for other DNA mimics like phage λ Gam protein (Wilkinson et al., 2016) and the recently discovered anti-CRISPR proteins (Wang et al., 2018). How the multifunctionality of these proteins are utilised by phage is an interesting topic for further study.

Comparisons with other RNAP inhibiting bacteriophage proteins

Several other bacteriophage proteins inhibit transcription. These include T7 gp2 and gp5.7, which inhibits E. coli RNAP, and P23-45 gp39 and gp76, which inhibit T. thermophilus transcription. Structural and biochemical studies show that gp2 specifically inhibits σ⁷⁰-dependent transcription through the interactions and stabilisation of the position of the inhibitory region 1.1 of σ⁷⁰, between the β and β’ clamps at the rim of the downstream DNA binding channel (Figure 7B; Bae et al., 2013). Gp39 binds at region 4 of σ^A and restricts its conformation so that it is unable to bind to −35 promoter regions (Figure 7C; Tagami et al., 2014). Gp76 on the other hand, binds deeply in the RNAP cleft, preventing open complex formation as well as transcription elongation (Figure 7D; Ooi et al., 2018). However, the binding site of gp76 is adjacent to, but not overlapping the σ⁷⁰ region σ_3.2, thus gp76 does not inhibit holoenzyme formation (Figure 7D).

In all the cases above, the bacteriophage proteins do not inhibit holoenzyme formation. Ocr on the other hand, inhibits holoenzyme formation per se and can disrupt a pre-formed holoenzyme (Figure 6A–B, Figure 6—figure supplement 1, Figure 7A). In addition, Ocr can also inhibit open complex formation but is unable to effectively disrupt pre-formed open complex (Figure 6).

Unique features of Ocr defines its ability in specific DNA processing systems

Ocr was shown to mimic a dsDNA in both shape and charge distribution (Walkinshaw et al., 2002) and an Ocr dimer mimics a slightly bent B-form DNA. Ocr has been shown to occupy the DNA binding grooves of type I RM enzymes through specific interactions, thus preventing viral DNA from being degraded or modified by the host RM system (Atanasiu et al., 2002; Kennaway et al., 2009).

The structural model of the archetypal Type I RM enzyme EcoKI (Kennaway et al., 2009) shows a close association of Ocr with the M.EcoKI modification methyltransferase core of the Type I RM enzyme in that Ocr is completely encircled by the enzyme, similar to DNA encircled by a DNA clamp (Kennaway et al., 2012; Walkinshaw et al., 2002) (PDB 1S7Z). Ocr, on the other hand, binds in the cleft, surrounded on three sides by RNAP. Therefore it is not surprising that both Ocr-RNAP and Ocr-M.EcoKI complexes are sufficiently stable to persist during lengthy size exclusion chromatography experiments (Figure 1 and Atanasiu et al., 2002).

In this work, we show that Ocr remains as a dimer and binds at the downstream DNA channel of RNAP via the complementary charge interactions between the positively charged RNAP channel and the negatively charged Ocr dimer. However, due to the extensive positively charged surface within RNAP, the binding is not specific. Indeed, two distinct binding modes have been detected which allow either the proximal or distal Ocr monomer to interact with the positively charged β’ clamp. The RNAP clamp opens up in the ‘wide clamp’ binding mode in order to accommodate the Ocr dimer conformation while maintaining the charge complementarity, both deep inside the cleft and with the β’ clamp. When the clamp is closed as in the transcriptional open complex, even though there is ample space to accommodate one Ocr subunit, the charge repulsion between the β’ jaw domain and the Ocr surface results in the opening of the β’ clamp, thus shifting Ocr upwards and away from the β’ jaw domain. These structures demonstrate that the structural flexibility of RNAP clamp plays a key role in its ability to accommodate Ocr. A study of the kinetics of the formation of the complexes between Ocr and RNAP would be interesting and may show multiple consecutive reactions as found for the interaction of a small DNA mimic Uracil glycosylase inhibitor (UGI) with uracil glycosylase (Bennett et al., 1993).

The interactions and structures of Ocr with RNAP observed here suggest that Ocr could potentially interact and inhibit other DNA processing enzymes through a series of non-specific interactions with dsDNA binding sites/channels. However, the rigidity and the bent conformation of the Ocr dimer imposes constraints on the binding sites, which are optimised for the RM enzymes it inhibits but can still be accommodated by the structurally flexible RNAP DNA binding channel. The rigidity of the dimer is defined by the interactions between the Ocr subunits. Work presented here suggests that Ocr could be modified to fine-tune its DNA mimicry for binding to specific DNA processing proteins, either by reducing the rigidity of the dimer or changing the bend angle to create a new dimer interface. DNA mimicry has been proposed as a potential effective therapeutic tool (Putnam and Tainer, 2005; Dryden, 2006; Roberts et al., 2012), Ocr could thus be exploited to create specific variants that can target specific DNA binding proteins, especially those that are shown to be antibiotic targets. In addition, our data here show that Ocr can effectively inhibit transcription under stress conditions when transcriptional activity is reduced. It is thus possible that during other growth conditions/phases, such as stationary phases or persistence, Ocr could be effectively utilised to inhibit transcription and thus diminish bacterial survival.

Materials and methods

RNAP-Ocr complex formation

E. coli RNA polymerase was purified as previously reported (Yang et al., 2015). As well as being purified as previously described (Sturrock et al., 2001), Ocr was ligated into pOPINF vector with a N-terminal His-tag. Three purification steps (Ni-NTA affinity, Heparin, and Gel filtration chromatography) were employed, which generated homogeneous Ocr protein as judged by SDS-PAGE gel. The RNAP-Ocr complex was purified by gel filtration chromatography. Initially RNAP (final concentration 14.5 µM) and Ocr (all Ocr concentrations are for the dimeric form, final concentration 58.0 µM) were mixed at a 1:4 molar ratio in a 150 µl reaction volume and the mixture (150 µl) was incubated on ice for 30 mins before being loaded onto the gel filtration column with a flowrate of 0.3 ml/min (Superose 6 10/300, GE Healthcare) and eluted in buffer containing 20 mM Tris pH8.0, 150 mM NaCl, 1 mM TCEP, 5 mM MgCl₂. The fractions were run on SDS-PAGE and the presence of Ocr was confirmed by Western blot against the His-tag. The same amount and concentration of RNAP and Ocr alone were also loaded on a Superose 6 10/300 column separately as a control .

Microscale thermophoresis (MST)

All MST experiments were performed using a Monolith NT.115 instrument (NanoTemper Technologies, Germany) at 22 °C. PBS (137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 1.8 mM KH₂PO₄) supplemented with 0.05% Tween-20 was used as MST-binding buffer for all experiments. In all cases hydrophobic treated capillaries were used. σ⁵⁴, σ^54-_R336A and σ⁷⁰ were purified to homogeneity similarly as described previously (Nechaev and Severinov, 1999; Yang et al., 2015). σ^54-_R336A and σ⁷⁰ holoenyzme were individually produced with gel filtration chromatography by mixing RNAP and σ with 1:4 molar ratio and incubating on ice for half an hour before loading to superpose 6 10/300 column.

Ocr, σ⁵⁴, and σ⁷⁰, σ^54-_R336A -RNAP holoenzyme and σ⁷⁰-RNAP holoenzyme were all His tagged and labelled with the kit (Monolith His-Tag Labeling Kit RED-tris-NTA 2nd Generation) separately. The labelled Ocr and sigma factors were diluted to a concentration of 50 nM and mixed with an equal volume of a serial dilution series of RNAP incubated at room temperature for 20 min before loading into MST capillaries thus giving a final concentration of 25 nM for Ocr and the sigma factors in the direct titration of RNAP. For Ocr competitive binding with σ⁷⁰ holoenzyme, 50 nM of his-labeled σ⁷⁰ holoenzyme, diluted in MST-binding buffer, was mixed with equal volumes of a dilution series of Ocr in MST-binding buffer. For DNA with holoenzyme binding affinity, linear promoter DNA (Zuo and Steitz, 2015) was used σ⁷⁰ and a nifH promoter with mismatch DNA between -10 and -1 (Glyde et al., 2018) was used. 40 nm of his-tagged holoenzyme was mixed with equal volumes of a dilution series of DNA in MST-binding buffer. MST experiments were performed using 70–100% LED power and 40–60% MST power with a wait time of 5 s, laser on time of 30 s and a back-diffusion time of 5 s. For all the binding experiments (Ocr or σ with RNAP, or DNA with holoenzyme), MST data were analysed using Grafit 3.0 (Erithacus Software) and the data were fitted with a quadratic tight binding equation with the concentration of the constituent being titrated kept constant in the fitting process. The dissociation constants obtained are maximum values and the real values are likely to be rather smaller. This is due to the concentration of the protein being titrated being larger than the dissociation constant. In the competitive binding experiment with Ocr and σ⁷⁰-RNAP holoenzyme, the data were fitted using the tight binding equation to determine the apparent dissociation constant. Each experiment was repeated at least three times.

CryoEM sample preparation

For structural studies using cryoEM, the His-tag of Ocr was cleaved and the purification procedure was as previously reported (Sturrock et al., 2001). The complex was formed as described above. The RNAP-Ocr complex sample was concentrated to 0.6 mg/ml and 3.5 µl of the complex was applied to R2/2 holey carbon grids (Quantifoil). The vitrification process was performed using a Vitrobot Mark IV (FEI) at 4°C and 95% humidity, blotted for 1.5 s with blotting force −6. The grid was then flash frozen in liquid ethane and stored in the liquid nitrogen before data collection.

Electron microscopy data collection

The cryoEM data were collected at eBIC (Diamond Light Source, UK) on a Titan Krios using EPU (FEI) operated at 300 kV and a K2 summit direct electron detector (Gatan). The data were collected with a defocus (underfocus) range between 1.5 µm to 3.5 µm. A total of 3543 micrographs were collected at pixel size of 1.06 Å/pixel and a dose of 50 e^-/Å² and each micrograph was fractioned into 41 frames (1.21 e^-/Å²/frame).

Image processing

The procedure of data processing is summarised in Figure 2—figure supplement 1. Frame alignment and dose weighting were carried out with MotionCor2 (Zheng et al., 2017) before estimating CTF parameters using Gctf (Zhang, 2016) and particle picking using Gautomatch (https://www.mrc-lmb.cam.ac.uk/kzhang/Gautomatch/) without a template. The picked particles were then extracted into boxes of 256 × 256 pixels. Initial 2D classification of the data was carried out in Cryosparc (Punjani et al., 2017) to remove junk particles due to ice contamination or other defects on the grids. Subsequent image processing was carried out in Relion 2.1 (Scheres, 2012). Briefly, the particles were separated using 3D classification. Three out of five classes were subsequently refined using the RNAP from the closed complex RPc (EMD-3695) filtered to 60 Å as the initial reference map. The remaining two classes were discarded due to their lack of clear structural features, probably derived from particles that are of poor quality. After combining the remaining three classes and refinement, one more round of focused 3D classification by applying a mask was carried out to separate different complexes or conformations. To generate the mask, structural models of RNAP and Ocr were first fitted into the refined 3D reconstruction. Ocr and the surrounding RNAP clamp regions were then used to calculate a volume in Chimera. Two classes with clearly different conformations of RNAP and both with Ocr bound, were refined, polished and post-processed (with masking and sharpening), resulting in the final reconstructions at 3.7 Å (for the wide clamp RNAP-Ocr) and 3.8 Å (for the narrow clamp RNAP-Ocr) based on the gold-standard Fourier shell correlation (FSC) at 0.143 criterion. The directional FSC analysis and histogram of directional resolution were performed using the 3D-FSC server (3dfsc.salk.edu) (Tan et al., 2017).

Model building, refinement and structural analysis

The RNAP from RPo (PDB code: 6GH5) and RPip (PDB code: 6GH6) was used as an initial model for the model building of narrow clamp and wide clamp reconstructions, respectively. Briefly, the RNAP was first fitted into the RNAP-Ocr density map in Chimera (Goddard et al., 2007). Subsequently the RNAP structure was subject to flexible fitting using MDFF (Trabuco et al., 2009). The Ocr crystal structure (PDB code: 1S7Z) was manually fitted into the extra density of the RNAP-Ocr map in Coot (Emsley et al., 2010). Jelly body refinement in Refmac (Murshudov et al., 2011) and real space refinement in Phenix (Afonine et al., 2012) were used to improve the model quality. The final statistics of the model are in Table 1. The figures used for structure analysis and comparison were produced in Pymol (The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC) and UCSF Chimera (Goddard et al., 2007).

Table 1. Cryo-EM data collection and refinement statistics.

	RNAP-Ocr
Data collection and processing
Magnification	47393
Total micrographs	3543
Movie frames	41
Pixel size (Å)	1.055
Defocus range (µm)	−1.2 to −3
Voltage (kV)	300
Electron dose (e^-/Å⁻²)	49.53
Total particles	753783
FSC threshold	0.143
Reconstruction (RELION)	Wide clamp	Narrow clamp
Particles	33646	27312
Resolution (Å)	3.7	3.8
Refinement
Resolution (Å)	3.7	3.8
R.m.s. deviations
Bond length (Å)	0.003	0.003
Bond angle (°)	0.816	0.865
Ramachandran plot
Favored regions (%)	91.88	91.05
Allowed regions (%)	8.06	8.69
Outlier	0.06	0.27
Validation
All-atom clashscore	4.63	4.5
Rotamer outliers (%)	0.15	0.18
C-beta deviations	0	0

Open in a new tab

Native gel mobility assay and western blotting

σ⁵⁴ as well as σ⁷⁰ were purified to homogeneity as described previously (Nechaev and Severinov, 1999; Yang et al., 2015). All the reactions were carried out in the binding buffer (20 mM Tris-HCl pH8.0, 200 mM KCl, 5 mM MgCl_2,1 mM DTT, 5% glycerol) at room temperature, for the sequentially addition of different components in each step, the incubation time was approximately15-20 min and all the gels used were polyacrylamide. To test the effects of Ocr on holoenzyme formation for both σ⁵⁴ and σ⁷⁰, all the reactions were carried out in 20 µl at room temperature. The order of addition, final concentrations and molar ratio of different components are listed in corresponding figures. For each 20 µl reaction, 10 µl was taken out and loaded into one polyacrylamide gel (4.5%) and the remaining 10 µl of each sample was loaded to another identical gel (4.5%) and run in parallel for 160 mins at 80 volts at 4°C cold room. One gel was then stained with Coomassie blue and visualised, whilst the other gel was used for western blotting against the His-tagged protein of either σ⁵⁴ or σ⁷⁰, following the standard western blotting protocol (iBlot 2 Dry Blotting System, Thermo Fisher Scientific).

To test the effects of Ocr on the formation σ⁷⁰ open complexes, DNA probes used in this study were all Cy3 labelled. For σ⁷⁰ open complex experiments, one linear DNA used, the non-template DNA used was 5’- ACTTGACATCCCACCTCACGTATGCTATAATCCTACGAGTCTGACGCGG −3’, and the template DNA used was 5’-CCGCGTCAGACTCGTAGGATTATAGCATACGTGAGGTGGGATGTCAAGT −3’. All the other reactions were carried out at room temperature. The lacUV5 linear DNA used was the same as reported (Cámara et al., 2010). The sequence of addition, the final concentration, molar ratios of different reactions are described in respective figures. All the gel (5%) are run at 4 °C in cold rooms, with 100 volts for 75 mins. The results were analysed by visualizing and quantifying the fluorescent signals of Cy3, attached to the DNA probes.

All the experiments were repeated at least once and the results are consistent with each other.

Small primed RNA assays

All reactions were performed in 10 µl final volumes containing: STA buffer (Burrows et al., 2010), 100 nM holoenzyme (1:4 ratio of RNAP:σ⁵⁴), 20 nM promoter DNA probe (for σ⁵⁴ dependent transcription open complex formation, 4 mM dATP and 5 mM PspF_1-275 were also present) and incubated at room temperature. The sp RNA synthesis was initiated by adding 0.5 mM dinucleotide primer (ApG, ApA, CpA or UpG), 0.2 mCi/ml [α-³²P] GTP (3000 Ci/mmol) or 0.2 mCi/ml [α-³²P] UTP (3000 Ci/mmol). The reaction mixtures were quenched by addition of 4 µl of denaturing formamide loading buffer and run on a 20% denaturing gel and visualised using a Fuji FLA-5000 Phosphorimager. At least three independent experiments were carried out and values were within 5% of the relative % values measured.

Bacterial growth curves

Growth of E. coli MG1655 cells not expressing Ocr (WT/pBAD18cm) and expressing Ocr (WT/pBAD18cm[ocr]) was monitored for 20 hr using a FLUOstar Omega plate reader (BMG LABTECH). Starting cultures of 0.1 OD₆₀₀ were inoculated in four different media, Luria-Bertani Broth (LB) with low (5 g/L) and high (10 g/L) salt concentration, nutrient broth (Oxoid) and modified M9 medium (Teknova), at two different temperatures, 37°C and 25°C. Expression of Ocr was induced using two different concentration of arabinose, 0.02%(^w/_v) and 0.2%(^w/_v).

Beta-galactosidase assays

Gene expression levels of beta-galactosidase in E. coli MG1655 cells not expressing Ocr (WT/pBAD18cm) and expressing Ocr (WT/pBAD18cm[ocr]) were assessed using 0.002 mg/ml fluorescein di(beta-D-galactopyranoside) (FDG; Sigma-Aldrich). FDG results in fluorescence signals, which are proportional to the enzymatic activities after being hydrolysed to fluorescein by beta-galactosidase. Beta-galactosidase expression was induced by 1 mM IPTG 5 hr post-inoculation.

RNA extraction, DNase digestion, reverse transcription and quantitative polymerase chain reaction (RT-qPCR)

Following the beta-galactosidase assays, total bacterial RNA was extracted from E. coli MG1655 cells using the RNeasy Mini Kit (Qiagen) according to the manufacturer’s instructions. Residual DNA was digested by DNase I (Promega) and cDNA synthesis was performed using 500 ng of RNA and the SuperScript IV reverse transcriptase (Invitrogen) according to the manufacturer’s instructions. Quantitation of the beta-galactosidase mRNA levels was performed using the specific oligonucleotide primers 5’-ATG GGT AAC AGT CTT GGC GG-3’ and 5’-GGC GTA TCG CCA AAA TCA CC-3’, the Power SYBR Green PCR Master Mix (Applied Biosystems) and the relative standard curve quantitation method as implemented by the OneStep Plus Real-Time qPCR System (Applied Biosystems).

Acknowledgements

Initial screening was carried out at Imperial College London Centre for Structural Biology EM facility. High resolution data were collected at the eBIC, Diamond Light Source (proposal EM14769). eBIC is funded by the Wellcome Trust, MRC and BBSRC. This work is funded by the BBSRC to XZ and MB (BB/N007816). XL is funded by a China Scholarship Council studentship. We are especially grateful to David Ratner for information concerning his experiments.

Funding Statement

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Contributor Information

Xiaodong Zhang, Email: xiaodong.zhang@imperial.ac.uk.

Seth Darst, The Rockefeller University, United States.

Cynthia Wolberger, Johns Hopkins University School of Medicine, United States.

Funding Information

This paper was supported by the following grants:

Biotechnology and Biological Sciences Research Council BB/N007816/1 to Fuzhou Ye, Milija Jovanovic, Martin Buck, Xiaodong Zhang.
China Scholarship Council Studentship to Xiaojiao Liu.

Additional information

Competing interests

No competing interests declared.

Author contributions

Data curation, Formal analysis, Validation, Visualization.

Data curation, Formal analysis, Validation, Investigation, Visualization.

Data curation.

Conceptualization, Resources, Methodology.

Resources, Supervision, Funding acquisition.

Conceptualization, Formal analysis, Supervision, Funding acquisition, Project administration.

Additional files

Supplementary file 1. Pairwise comparisons of E. coli cells expressing Ocr (WT/pBAD18cm[ocr]) versus cells not expressing Ocr (WT/pBAD18cm).

elife-52125-supp1.docx^{(14.4KB, docx)}

Transparent reporting form

elife-52125-transrepform.docx^{(245.8KB, docx)}

Data availability

All data generated or analysed during the study are included in the manuscript and supporting files. The cryo EM maps and structural models will be deposited into EMDB and PDB with access codes 6R9G and 6R9B.

The following datasets were generated:

Ye FZ, Zhang XD. 2020. Structural basis of transcription inhibition by the DNA mimic Ocr protein of bacteriophage T7. RCSB Protein Data Bank. 6R9G

Ye FZ, Zhang XD. 2020. Structural basis of transcription inhibition by the DNA mimic Ocr protein of bacteriophage T7. RCSB Protein Data Bank. 6R9B

References

Afonine PV, Grosse-Kunstleve RW, Echols N, Headd JJ, Moriarty NW, Mustyakimov M, Terwilliger TC, Urzhumtsev A, Zwart PH, Adams PD. Towards automated crystallographic structure refinement with phenix.refine. Acta Crystallographica Section D Biological Crystallography. 2012;68:352–367. doi: 10.1107/S0907444912001308. [DOI] [PMC free article] [PubMed] [Google Scholar]
Atanasiu C, Su TJ, Sturrock SS, Dryden DT. Interaction of the ocr gene 0.3 protein of bacteriophage T7 with EcoKI restriction/modification enzyme. Nucleic Acids Research. 2002;30:3936–3944. doi: 10.1093/nar/gkf518. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bae B, Davis E, Brown D, Campbell EA, Wigneshweraraj S, Darst SA. Phage T7 Gp2 inhibition of Escherichia coli RNA polymerase involves misappropriation of σ70 domain 1.1. PNAS. 2013;110:19772–19777. doi: 10.1073/pnas.1314576110. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bae B, Feklistov A, Lass-Napiorkowska A, Landick R, Darst SA. Structure of a bacterial RNA polymerase holoenzyme open promoter complex. eLife. 2015;4:e08504. doi: 10.7554/eLife.08504. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bandyopadhyay PK, Studier FW, Hamilton DL, Yuan R. Inhibition of the type I restriction-modification enzymes EcoB and EcoK by the gene 0.3 protein of bacteriophage T7. Journal of Molecular Biology. 1985;182:567–578. doi: 10.1016/0022-2836(85)90242-6. [DOI] [PubMed] [Google Scholar]
Bennett SE, Schimerlik MI, Mosbaugh DW. Kinetics of the uracil-DNA glycosylase/inhibitor protein association ung interaction with ugi, nucleic acids, and uracil compounds. The Journal of Biological Chemistry. 1993;268:26879–26885. [PubMed] [Google Scholar]
Browning DF, Busby SJ. The regulation of bacterial transcription initiation. Nature Reviews Microbiology. 2004;2:57–65. doi: 10.1038/nrmicro787. [DOI] [PubMed] [Google Scholar]
Browning DF, Busby SJ. Local and global regulation of transcription initiation in Bacteria. Nature Reviews Microbiology. 2016;14:638–650. doi: 10.1038/nrmicro.2016.103. [DOI] [PubMed] [Google Scholar]
Burrows PC, Joly N, Buck M. A prehydrolysis state of an AAA+ ATPase supports transcription activation of an enhancer-dependent RNA polymerase. PNAS. 2010;107:9376–9381. doi: 10.1073/pnas.1001188107. [DOI] [PMC free article] [PubMed] [Google Scholar]
Cámara B, Liu M, Reynolds J, Shadrin A, Liu B, Kwok K, Simpson P, Weinzierl R, Severinov K, Cota E, Matthews S, Wigneshweraraj SR. T7 phage protein Gp2 inhibits the Escherichia coli RNA polymerase by antagonizing stable DNA strand separation near the transcription start site. PNAS. 2010;107:2247–2252. doi: 10.1073/pnas.0907908107. [DOI] [PMC free article] [PubMed] [Google Scholar]
Chakraborty A, Wang D, Ebright YW, Korlann Y, Kortkhonjia E, Kim T, Chowdhury S, Wigneshweraraj S, Irschik H, Jansen R, Nixon BT, Knight J, Weiss S, Ebright RH. Opening and closing of the bacterial RNA polymerase clamp. Science. 2012;337:591–595. doi: 10.1126/science.1218716. [DOI] [PMC free article] [PubMed] [Google Scholar]
Chaney M, Buck M. The sigma 54 DNA-binding domain includes a determinant of enhancer responsiveness. Molecular Microbiology. 1999;33:1200–1209. doi: 10.1046/j.1365-2958.1999.01566.x. [DOI] [PubMed] [Google Scholar]
Decker KB, Hinton DM. Transcription regulation at the core: similarities among bacterial, archaeal, and eukaryotic RNA polymerases. Annual Review of Microbiology. 2013;67:113–139. doi: 10.1146/annurev-micro-092412-155756. [DOI] [PubMed] [Google Scholar]
Dryden DT. DNA mimicry by proteins and the control of enzymatic activity on DNA. Trends in Biotechnology. 2006;24:378–382. doi: 10.1016/j.tibtech.2006.06.004. [DOI] [PubMed] [Google Scholar]
Emsley P, Lohkamp B, Scott WG, Cowtan K. Features and development of coot. Acta Crystallographica. Section D, Biological Crystallography. 2010;66:486–501. doi: 10.1107/S0907444910007493. [DOI] [PMC free article] [PubMed] [Google Scholar]
Feklístov A, Sharon BD, Darst SA, Gross CA. Bacterial sigma factors: a historical, structural, and genomic perspective. Annual Review of Microbiology. 2014;68:357–376. doi: 10.1146/annurev-micro-092412-155737. [DOI] [PubMed] [Google Scholar]
Glyde R, Ye F, Jovanovic M, Kotta-Loizou I, Buck M, Zhang X. Structures of bacterial RNA polymerase complexes reveal the mechanism of DNA loading and transcription initiation. Molecular Cell. 2018;70:1111–1120. doi: 10.1016/j.molcel.2018.05.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
Goddard TD, Huang CC, Ferrin TE. Visualizing density maps with UCSF chimera. Journal of Structural Biology. 2007;157:281–287. doi: 10.1016/j.jsb.2006.06.010. [DOI] [PubMed] [Google Scholar]
Hahn S, Young ET. Transcriptional regulation in Saccharomyces cerevisiae: transcription factor regulation and function, mechanisms of initiation, and roles of activators and coactivators. Genetics. 2011;189:705–736. doi: 10.1534/genetics.111.127019. [DOI] [PMC free article] [PubMed] [Google Scholar]
Hausmann R, Messerschmid M. Inhibition of gene expression of T7-related phages by prophage P1. Molecular and General Genetics MGG. 1988;212:543–547. doi: 10.1007/BF00330862. [DOI] [PubMed] [Google Scholar]
Ho MX, Hudson BP, Das K, Arnold E, Ebright RH. Structures of RNA polymerase–antibiotic complexes. Current Opinion in Structural Biology. 2009;19:715–723. doi: 10.1016/j.sbi.2009.10.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ho CH, Wang HC, Ko TP, Chang YC, Wang AH. The T4 phage DNA mimic protein arn inhibits the DNA binding activity of the bacterial Histone-like protein H-NS. The Journal of Biological Chemistry. 2014;289:27046–27054. doi: 10.1074/jbc.M114.590851. [DOI] [PMC free article] [PubMed] [Google Scholar]
Issinger OG, Hausmann R. Synthesis of phage coded products during infection of E. coli in gene-1-mutants of phages T3 and T7. Hoppe-Seyler's Zeitschrift Fur Physiologische Chemie. 1972;353:1530. [PubMed] [Google Scholar]
Kelleher JE, Raleigh EA. Response to UV damage by four Escherichia coli K-12 restriction systems. Journal of Bacteriology. 1994;176:5888–5896. doi: 10.1128/JB.176.19.5888-5896.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kennaway CK, Obarska-Kosinska A, White JH, Tuszynska I, Cooper LP, Bujnicki JM, Trinick J, Dryden DT. The structure of m.ecoki type I DNA methyltransferase with a DNA mimic antirestriction protein. Nucleic Acids Research. 2009;37:762–770. doi: 10.1093/nar/gkn988. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kennaway CK, Taylor JE, Song CF, Potrzebowski W, Nicholson W, White JH, Swiderska A, Obarska-Kosinska A, Callow P, Cooper LP, Roberts GA, Artero JB, Bujnicki JM, Trinick J, Kneale GG, Dryden DT. Structure and operation of the DNA-translocating type I DNA restriction enzymes. Genes & Development. 2012;26:92–104. doi: 10.1101/gad.179085.111. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kornberg RD. Mechanism and regulation of yeast RNA polymerase II transcription. Cold Spring Harbor. 1998;63:229–232. doi: 10.1101/sqb.1998.63.229. [DOI] [PubMed] [Google Scholar]
Krüger DH, Schroeder C. Bacteriophage T3 and bacteriophage T7 virus-host cell interactions. Microbiological Reviews. 1981;45:9–51. doi: 10.1128/MMBR.45.1.9-51.1981. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ma C, Yang X, Lewis PJ. Bacterial transcription as a target for antibacterial drug development. Microbiology and Molecular Biology Reviews. 2016;80:139–160. doi: 10.1128/MMBR.00055-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
McAllister WT, Barrett CL. Roles of the early genes of bacteriophage T7 in shutoff of host macromolecular synthesis. Journal of Virology. 1977;23:543–553. doi: 10.1128/JVI.23.3.543-553.1977. [DOI] [PMC free article] [PubMed] [Google Scholar]
Murshudov GN, Skubák P, Lebedev AA, Pannu NS, Steiner RA, Nicholls RA, Winn MD, Long F, Vagin AA. REFMAC5 for the refinement of macromolecular crystal structures. Acta Crystallographica Section D Biological Crystallography. 2011;67:355–367. doi: 10.1107/S0907444911001314. [DOI] [PMC free article] [PubMed] [Google Scholar]
Nechaev S, Severinov K. Inhibition of Escherichia coli RNA polymerase by bacteriophage T7 gene 2 protein. Journal of Molecular Biology. 1999;289:815–826. doi: 10.1006/jmbi.1999.2782. [DOI] [PubMed] [Google Scholar]
Ooi WY, Murayama Y, Mekler V, Minakhin L, Severinov K, Yokoyama S, Sekine SI. A Thermus phage protein inhibits host RNA polymerase by preventing template DNA strand loading during open promoter complex formation. Nucleic Acids Research. 2018;46:431–441. doi: 10.1093/nar/gkx1162. [DOI] [PMC free article] [PubMed] [Google Scholar]
Punjani A, Rubinstein JL, Fleet DJ, Brubaker MA. cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nature Methods. 2017;14:290–296. doi: 10.1038/nmeth.4169. [DOI] [PubMed] [Google Scholar]
Putnam CD, Tainer JA. Protein mimicry of DNA and pathway regulation. DNA Repair. 2005;4:1410–1420. doi: 10.1016/j.dnarep.2005.08.007. [DOI] [PubMed] [Google Scholar]
Ratner D. The interaction bacterial and phage proteins with immobilized Escherichia coli RNA polymerase. Journal of Molecular Biology. 1974;88:373–383. doi: 10.1016/0022-2836(74)90488-4. [DOI] [PubMed] [Google Scholar]
Roberts GA, Stephanou AS, Kanwar N, Dawson A, Cooper LP, Chen K, Nutley M, Cooper A, Blakely GW, Dryden DT. Exploring the DNA mimicry of the ocr protein of phage T7. Nucleic Acids Research. 2012;40:8129–8143. doi: 10.1093/nar/gks516. [DOI] [PMC free article] [PubMed] [Google Scholar]
Scheres SH. RELION: implementation of a bayesian approach to cryo-EM structure determination. Journal of Structural Biology. 2012;180:519–530. doi: 10.1016/j.jsb.2012.09.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
Srivastava A, Talaue M, Liu S, Degen D, Ebright RY, Sineva E, Chakraborty A, Druzhinin SY, Chatterjee S, Mukhopadhyay J, Ebright YW, Zozula A, Shen J, Sengupta S, Niedfeldt RR, Xin C, Kaneko T, Irschik H, Jansen R, Donadio S, Connell N, Ebright RH. New target for inhibition of bacterial RNA polymerase: 'switch region'. Current Opinion in Microbiology. 2011;14:532–543. doi: 10.1016/j.mib.2011.07.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
Studier FW. Bacteriophage T7. Science. 1972;176:367–376. doi: 10.1126/science.176.4033.367. [DOI] [PubMed] [Google Scholar]
Studier FW. Gene 0.3 of bacteriophage T7 acts to overcome the DNA restriction system of the host. Journal of Molecular Biology. 1975;94:283–295. doi: 10.1016/0022-2836(75)90083-2. [DOI] [PubMed] [Google Scholar]
Sturrock SS, Dryden DT, Atanasiu C, Dornan J, Bruce S, Cronshaw A, Taylor P, Walkinshaw MD. Crystallization and preliminary X-ray analysis of ocr, the product of gene 0.3 of bacteriophage T7. Acta Crystallographica Section D Biological Crystallography. 2001;57:1652–1654. doi: 10.1107/S0907444901011623. [DOI] [PubMed] [Google Scholar]
Tabib-Salazar A, Liu B, Shadrin A, Burchell L, Wang Z, Wang Z, Goren MG, Yosef I, Qimron U, Severinov K, Matthews SJ, Wigneshweraraj S. Full shut-off of Escherichia coli RNA-polymerase by T7 phage requires a small phage-encoded DNA-binding protein. Nucleic Acids Research. 2017;45:7697–7707. doi: 10.1093/nar/gkx370. [DOI] [PMC free article] [PubMed] [Google Scholar]
Tabib-Salazar A, Liu B, Barker D, Burchell L, Qimron U, Matthews SJ, Wigneshweraraj S. T7 phage factor required for managing RpoS in Escherichia coli. PNAS. 2018;115:E5353–E5362. doi: 10.1073/pnas.1800429115. [DOI] [PMC free article] [PubMed] [Google Scholar]
Tagami S, Sekine S, Minakhin L, Esyunina D, Akasaka R, Shirouzu M, Kulbachinskiy A, Severinov K, Yokoyama S. Structural basis for promoter specificity switching of RNA polymerase by a phage factor. Genes & Development. 2014;28:521–531. doi: 10.1101/gad.233916.113. [DOI] [PMC free article] [PubMed] [Google Scholar]
Tan YZ, Baldwin PR, Davis JH, Williamson JR, Potter CS, Carragher B, Lyumkis D. Addressing preferred specimen orientation in single-particle cryo-EM through tilting. Nature Methods. 2017;14:793–796. doi: 10.1038/nmeth.4347. [DOI] [PMC free article] [PubMed] [Google Scholar]
Trabuco LG, Villa E, Schreiner E, Harrison CB, Schulten K. Molecular dynamics flexible fitting: a practical guide to combine cryo-electron microscopy and X-ray crystallography. Methods. 2009;49:174–180. doi: 10.1016/j.ymeth.2009.04.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
Vassylyev DG, Vassylyeva MN, Perederina A, Tahirov TH, Artsimovitch I. Structural basis for transcription elongation by bacterial RNA polymerase. Nature. 2007;448:157–162. doi: 10.1038/nature05932. [DOI] [PubMed] [Google Scholar]
Walkinshaw MD, Taylor P, Sturrock SS, Atanasiu C, Berge T, Henderson RM, Edwardson JM, Dryden DT. Structure of ocr from bacteriophage T7, a protein that mimics B-form DNA. Molecular Cell. 2002;9:187–194. doi: 10.1016/S1097-2765(02)00435-5. [DOI] [PubMed] [Google Scholar]
Wang HC, Chou CC, Hsu KC, Lee CH, Wang AH. New paradigm of functional regulation by DNA mimic proteins: recent updates. IUBMB Life. 2018;72:539–548. doi: 10.1002/iub.1992. [DOI] [PubMed] [Google Scholar]
Wilkinson M, Troman L, Wan Nur Ismah WA, Chaban Y, Avison MB, Dillingham MS, Wigley DB. Structural basis for the inhibition of RecBCD by gam and its synergistic antibacterial effect with quinolones. eLife. 2016;5:e22963. doi: 10.7554/eLife.22963. [DOI] [PMC free article] [PubMed] [Google Scholar]
Yang Y, Darbari VC, Zhang N, Lu D, Glyde R, Wang YP, Winkelman JT, Gourse RL, Murakami KS, Buck M, Zhang X. Transcription structures of the rna polymerase-σ54 reveal new and conserved regulatory strategies. Science. 2015;349:882–885. doi: 10.1126/science.aab1478. [DOI] [PMC free article] [PubMed] [Google Scholar]
Zhang Y, Feng Y, Chatterjee S, Tuske S, Ho MX, Arnold E, Ebright RH. Structural basis of transcription initiation. Science. 2012;338:1076–1080. doi: 10.1126/science.1227786. [DOI] [PMC free article] [PubMed] [Google Scholar]
Zhang K. Gctf: real-time CTF determination and correction. Journal of Structural Biology. 2016;193:1–12. doi: 10.1016/j.jsb.2015.11.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
Zheng SQ, Palovcak E, Armache J-P, Verba KA, Cheng Y, Agard DA. MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nature Methods. 2017;14:331–332. doi: 10.1038/nmeth.4193. [DOI] [PMC free article] [PubMed] [Google Scholar]
Zuo Y, Steitz TA. Crystal structures of the E. coli transcription initiation complexes with a complete bubble. Molecular Cell. 2015;58:534–540. doi: 10.1016/j.molcel.2015.03.010. [DOI] [PMC free article] [PubMed] [Google Scholar]

eLife. doi: 10.7554/eLife.52125.sa1

Decision letter

Editor: Seth Darst¹

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

The results in this manuscript confirm that bacteriophage T7-encoded Ocr interacts with host E. coli RNAP, show that Ocr inhibits transcription by host RNAP, indicate that Ocr inhibits transcription by host RNAP through disruption of RNAP-σ and RNAP-DNA interactions, and define structures of RNAP-Ocr complexes that show Ocr interacts with the RNAP DNA binding determinants as a molecular mimic of DNA and suggest Ocr disrupts RNAP-σ and RNAP-DNA interactions through steric clash. The findings provide a striking example of protein function through structural mimicry of DNA. In addition, the results may provide an example of dual-targeted regulation – namely, a bacteriophage regulator that may functionally inhibit both host type I restriction endonucleases and host RNAP.

Decision letter after peer review:

Thank you for submitting your article "Structural basis of transcription inhibition by the DNA mimic protein Ocr of bacteriophage T7" for consideration by eLife. Your article has been reviewed by two peer reviewers, and the evaluation has been overseen by a guest Reviewing Editor and Cynthia Wolberger as the Senior Editor.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

Previous studies showed that bacteriophage T7 gp0.3 (Ocr) overcomes host restriction by serving as a molecular mimic of double-stranded DNA that binds to, and inhibits, host type I restriction endonucleases. Early work had also suggested that Ocr may interact with host RNAP. The results in this manuscript confirm that Ocr interacts with host RNAP and show that Ocr inhibits transcription by host RNAP by disrupting interactions with both σ factor and DNA interactions. The authors determine structures of RNAP-Ocr complexes, which show that Ocr interacts with the RNAP DNA binding determinants by serving as a molecular mimic of DNA and suggest Ocr disrupts RNAP-σ and RNAP-DNA interactions through steric clash. The findings provide a striking example of protein function through structural mimicry of DNA. In addition, the results may provide an example of dual-targeted regulation – namely, a bacteriophage regulator that may functionally inhibit both host type I restriction endonucleases and host RNAP.

Overall the structural work itself is of significant interest but requires substantial revision in terms of presentation. The part of the manuscript on mechanism is weaker but could be improved substantially with revisions (detailed below). The main weakness of the manuscript is that it does not sufficiently address to what extent the observed interaction of Ocr with RNAP is biologically relevant for bacteriophage T7 virulence, and how its functional importance compares to Ocr's interaction with type I restriction enzymes. The reviewers suggest that additional experiments addressing functional relevance is within the scope of the manuscript and are desirable but not essential.

Essential revisions:

I) Presentation of structural studies:

1) A broad issue concerning the presentation of the structural results is the overall focus on comparing Ocr-RNAP complexes with σ⁵⁴-holoenzyme complexes since functional relevance of Ocr inhibition of σ⁵⁴-dependent transcription is extremely unlikely. The reviewers agree that it's important to show the relationship between binding modes of Ocr in the Ocr-RNAP structures and binding modes of DNA in transcription initiation complexes, but that should be done with σ⁷⁰-RNAP structures (which are all available). The presentation here makes it appear that phage-mediated inhibition of σ⁵⁴-dependent transcription is a central component of phage infection, which is almost certainly not the case. The figures and data concerning Ocr inhibition of σ⁵⁴-dependent transcription should be removed and/or moved to the supplement. These changes concern (although may not be limited to):

Figure 2C should show σ⁷⁰-RPo complex, not σ⁵⁴-RPo (6GH5).

Figure 4 should use a σ⁷⁰-RPo, not σ⁵⁴-RPo.

Data presented in Figure 6D, E – possibly irrelevant.

Data presented in Figure 7A, C – possibly irrelevant.

Data presented in Figure 8A – possibly irrelevant.

2) It is surprising that the authors did not compare the Ocr dimer structures in their cryo-EM maps with previously published Ocr structures, both in isolation and bound to EcoKI. Such a comparison, as well as of the interactions of Ocr with RNAP and EcoKI, is warranted at least as a comment in the main text, if not in a figure.

3) There are many issues regarding the figures and presentation of structural and mechanistic data. These are detailed below (Figure revisions).

II) Studies of Ocr inhibition mechanism:

1) The data on transcription inhibition in uninfected cells engineered to overproduce Ocr (Figure 5) is relatively uninformative and should be removed.

2) Figure 6, presenting confusing 'reaction schematics' and gel images for 44 different assays and conditions should be replaced by bar graphs with error bars, presenting the most important assays and conditions.

3) The in vitro transcription experiments of Figure 6 were performed with fully duplex promoter DNAs, while the gel shifts and competition experiments of Figure 7C and 7D are done with promoter DNAs that have mismatch transcription bubbles – that explains why Ocr reduces in vitro transcription from 100 to 39%, but in 7D (for example), there is no decrease in open complex formation whatsoever. It’s not clear why the experiments were done this way – the gel shift/competition experiments with mismatch promoters aren't really very informative of anything. These experiments should be removed or repeated with relevant DNA templates (i.e. fully duplex).

4) The text describing the Ocr inhibition mechanism could be shortened.

III) Biological relevance of inhibition of RNAP function by Ocr to T7 phage virulence:

The weakest part of the manuscript is the proposal that Ocr functionally inhibits host RNAP during infection by bacteriophage T7 – i.e., the proposal that inhibition of host RNAP by OCR occurs during, and is important for, infection by bacteriophage T7. This proposal should be either: (1) tempered with strong caveats, or (2) bolstered by additional experimental data. One experiment that potentially could be performed quickly would be to compare burst sizes for infection of host cells lacking type I restriction endonucleases (available commercially) with bacteriophage T7 derivatives having or lacking the gene encoding Ocr; If the authors proposal is true, there should be measurable differences in burst sizes. The reviewers strongly encourage, but do not absolutely require, the authors to attempt to compare burst sizes of ocr+ and ocr= T7 phage in hosts strains lacking type I restriction.

Figure Revisions:

Figure 1:

1) In A, there are two peaks of the RNAP-Ocr complex eluting between 40-57 ml. Do the authors know what the difference is between the complexes in both peaks? Which peak was used for structural studies?

2) In C, it is unclear why the RNAP-Ocr complex is loaded twice in the western blot. What's the difference between both lanes?

3) In D-E, the authors use EC50 and Kd interchangeably. To avoid confusion, the authors should be consistent and refer to Kd. Moreover, considering the labeled ligand in these experiments is at 25 nM concentration and the Kd for the interactions is in a similar concentration range, the binding measurements are performed under conditions of RNAP depletion (i.e. the free concentration of RNAP does not equal the concentration of added RNAP for each data point). If the authors cannot use ligand concentration 10-fold below the Kd for technical reasons, they should use a binding model that accounts for this change in free RNAP concentration or at least acknowledge that the reported Kds are higher than the true Kds.

4) The authors also state that Ocr is much more abundant in T7 infected cells than type I restriction enzymes, arguing that the additional Ocr molecules interact with PNAP. It would be helpful if the authors could comment on the intracellular concentration of Ocr compared to σ factors and whether these are in a range where Ocr can effectively compete with σ factors for binding to RNAP.

Figure 2:

5) Should be revised to compare Ocr-RNAP binding with σ⁷⁰-holoenzyme transcription initiation complexes.

6) For non-experts, it is not easily evident where the template and non-template DNA strands, as well as the RNA, are located in the structure. The authors should consider coloring these elements differently (also in Figure 3).

Figure 2—figure supplement 1:

7) The authors used focused 3D classification but do not explain how the masks were generated. A description should be included in the Materials and methods section, and it would be useful to also show the masks used in the figure.

Figure 2—figure supplement 2:

8) A and B, the particle orientation distributions are far from isotropic, the effect of this on the FSC curves should be assessed with directional FSC plots (https://3dfsc.salk.edu/). Some of the density maps shown in Figure 2—figure supplement 3 may suffer from this.

9) In A and B, the display threshold for rendering the EM volumes seems too low. Increasing the threshold will likely help to bring out the secondary structure features that should be easily identifiable at this resolution but are not well visible in the representations shown in the figures.

10) In D, what's the difference between both half-map FSCs? Presumably one map was used for model refinement and the other as control, but it is not clear which is which.

Figure 2—figure supplement 3:

11) The authors state that many side chain densities are resolved in the map, but this is not well visible in panel C.

Figure 4:

12) Should be revised to compare Ocr-RNAP binding with σ⁷⁰-holoenzyme transcription initiation complexes.

13) The authors state in the text (subsection “Structures of Ocr in complex with RNAP”, last paragraph) that the downstream DNA binding channel provides a positively charged surface composed of b, b' clamps and b' jaw, but this is not shown in the figure. The authors further state that positively charged areas on the Ocr surface face the b' jaw domain, but this is also not evident in this figure as the b' jaw is not shown in B. It would be helpful to revise the figure accordingly.

Figure 7 and Figure 7—figure supplement 1:

14) How were the gels stained in the different panels?

15) In Figure 7A and B, the authors argue that there is a gradual disappearance of the RNAP-σ⁵⁴ complex when adding increasing concentrations of Ocr. However, in conditions I and II in 7A there seems to be no or very little RNAP-σ⁵⁴ complex present at either Ocr concentration. A more careful titration would be necessary to make this point. The low abundance of RNAP-σ⁵⁴ in lanes 7 and 9 is also surprising since one would expect a similar abundance of both RNAP-Ocr and RNAP-σ⁵⁴ at equimolar concentration of the components given the similar Kds for the interactions. In condition III, the authors add Ocr to preformed RNAP-σ⁵⁴ purified by gel filtration, but it seems there is barely any σ⁵⁴ present (see also the very weak signal in Figure 7—figure supplement 1A when comparing lanes 11 and 12 to lane 6). Therefore, it is unclear whether Ocr actually competes with σ⁵⁴ or binds to abundant free RNAP. To substantiate their claim, the authors should demonstrate that the preformed RNAP-σ complexes contain stoichiometric amounts of all components as expected. More generally, the gels shown could also benefit from quantitation.

16) In Figure 7C and D, the authors changed the protein concentrations for the different reactions compared to A and B. To ease comparison with the other panels in this figure, it would make more sense to use similar experimental conditions as in the previous experiments. It would also be helpful if the authors could comment on the affinity of the DNA for RNAP-σ. Is binding tighter than that of Ocr to RNAP to explain that DNA can overcome the inhibitory effect of Ocr that is pre-bound to RNAP?

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Structural basis of transcription inhibition by the DNA mimic protein Ocr of bacteriophage T7" for further consideration by eLife. Your revised article has been evaluated by Cynthia Wolberger (Senior Editor) and a Reviewing Editor.

The manuscript has been improved and the comments of the reviewers have been satisfactorily addressed. However, there is a remaining issues with one of the figures that needs to be addressed before acceptance, as outlined below:

Figure 6 in the revised manuscript is problematic – the discussion of Figure 6 in the text, in the Figure 6 legend, and the figure itself are inconsistent. While Figure 6 contains panels A – D, the figure legend only refers to panels A – C. In addition, the legend does not adequately describe the figures. The legend to panel A should describe the experiment more completely and explain what Roman numerals I, II and III denote in that panel, as well as in panel B. It would perhaps be best to relabel the panels to denote the reaction scheme as panel A and to explicitly point the reader to that scheme in explaining I, II and III. The differences between the pairs of lanes 7-8, 9-10, 11-12 should also be explained.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for submitting your article "Structural basis of transcription inhibition by the DNA mimic protein Ocr of bacteriophage T7" for consideration by eLife. The evaluation has been overseen by a Reviewing Editor and Cynthia Wolberger as the Senior Editor.

Figure 6 and the legend are now improved in that one can better understand what is in the figure based on the legend. In panels C and D, the authors helpfully added labels above the gel lanes to explain the differences between the pairs of lanes in 'I', 'II', and 'III' (0.5 vs. 5 microM); however, the same needs to be done for panels A and B. Please label panels A and B with the concentrations of Ocr above the pairs of lanes in I, II, and III so that all the panels in the figure are labeled optimally and the same.

eLife. 2020 Feb 10;9:e52125. doi: 10.7554/eLife.52125.sa2

Author response

Essential revisions:

I) Presentation of structural studies:

1) A broad issue concerning the presentation of the structural results is the overall focus on comparing Ocr-RNAP complexes with σ⁵⁴-holoenzyme complexes since functional relevance of Ocr inhibition of σ⁵⁴-dependent transcription is extremely unlikely. The reviewers agree that it's important to show the relationship between binding modes of Ocr in the Ocr-RNAP structures and binding modes of DNA in transcription initiation complexes, but that should be done with σ⁷⁰-RNAP structures (which are all available). The presentation here makes it appear that phage-mediated inhibition of σ⁵⁴-dependent transcription is a central component of phage infection, which is almost certainly not the case. The figures and data concerning Ocr inhibition of σ⁵⁴-dependent transcription should be removed and/or moved to the supplement.

σ⁵⁴-dependent transcription plays vital roles in responding to phage infection. In fact, our model system for σ⁵⁴-dependent transcription is the psp (phage shock protein) system, which is induced by filamentous phage infection among others. However, we have no data in supporting whether σ⁵⁴-dependent transcription is a central component here and have replaced the figures with σ⁷⁰ and moved σ⁵⁴ comparisons, if relevant, to supplementary data.

These changes concern (although may not be limited to):

Figure 2C should show σ⁷⁰-RPo complex, not σ⁵⁴-RPo (6GH5).

We have done so.

Figure 4 should use a σ⁷⁰-RPo, not σ⁵⁴-RPo.

We have done so.

Data presented in Figure 6D, E – possibly irrelevant.

It is clear that Ocr inhibits both σ⁵⁴ and σ⁷⁰-dependent transcription in vitro although it is unclear the exact in vivo effect. We think the data should be presented but have moved σ⁵⁴-dependent data to supplementary.

Data presented in Figure 7A, C – possibly irrelevant.

Again we have decided to present the data but moved to supplementary.

Data presented in Figure 8A – possibly irrelevant.

We have removed the direct comparison with RNAP-σ⁵⁴ structure.

2) It is surprising that the authors did not compare the Ocr dimer structures in their cryo-EM maps with previously published Ocr structures, both in isolation and bound to EcoKI. Such a comparison, as well as of the interactions of Ocr with RNAP and EcoKI, is warranted at least as a comment in the main text, if not in a figure.

The resolution for Ocr is insufficient for building an atomic model so we used the Ocr crystal structure as the starting model and then fitted into the density and carried out refinement. Overall there are no significant differences at the monomeric level between the Ocr model in this study and that of crystal structure and the dimer interface is also largely unchanged. However the dimer is slightly less bent compared to that of crystal structure. We have added a sentence to describe this and a supplementary figure (Figure 2—figure supplement 4). In terms of EcoKI complex, the structural model was derived from a negatively stained reconstruction and the reliability and accuracy are questionable hence we are reluctant to perform a detailed comparison. Nevertheless, we have added a sentence stating that in that complex, Ocr is fully encircled by the enzyme, similar to DNA encircled by the DNA clamp while Ocr in the transcriptional complex is bound in the cleft.

3) There are many issues regarding the figures and presentation of structural and mechanistic data. These are detailed below (Figure revisions).

II) Studies of Ocr inhibition mechanism:

1) The data on transcription inhibition in uninfected cells engineered to overproduce Ocr (Figure 5) is relatively uninformative and should be removed.

Here we show that expression of Ocr in vivo inhibited transcription in a test reporter assay, consistent with in vitro data and suggesting that Ocr could perform as an antimicrobial. We think this is relevant information and have kept the main data in Figure 5 (Figure 5C-D) but reduced the text to describe these. We have also moved some of the data to new Figure 5—figure supplement 2.

2) Figure 6, presenting confusing 'reaction schematics' and gel images for 44 different assays and conditions should be replaced by bar graphs with error bars, presenting the most important assays and conditions.

We have added bar graphs as suggested and moved σ⁵⁴-related data to supplementary (new Figure 5—figure supplement 1).

3) The in vitro transcription experiments of Figure 6 were performed with fully duplex promoter DNAs, while the gel shifts and competition experiments of Figure 7C and D are done with promoter DNAs that have mismatch transcription bubbles – that explains why Ocr reduces in vitro transcription from 100 to 39%, but in 7D (for example), there is no decrease in open complex formation whatsoever. It's not clear why the experiments were done this way – the gel shift/competition experiments with mismatch promoters aren't really very informative of anything. These experiments should be removed or repeated with relevant DNA templates (i.e. fully duplex).

We have now repeated these experiments with fully duplex DNA and results are shown in new Figure 6 and have quantified and added bar graphs. We performed mismatched DNA for both σ⁵⁴ and σ⁷⁰ to compare effects on open complex, which turned out to be negligible. We have now removed these data. The competition experiments for promoter complex formation (Figure 6) are now under the same conditions as those of in vitro transcription assays (Figure 5) and indeed they agree very well with each other.

4) The text describing the Ocr inhibition mechanism could be shortened.

We have now shortened this part.

III) Biological relevance of inhibition of RNAP function by Ocr to T7 phage virulence:

The weakest part of the manuscript is the proposal that Ocr functionally inhibits host RNAP during infection by bacteriophage T7 – i.e., the proposal that inhibition of host RNAP by OCR occurs during, and is important for, infection by bacteriophage T7. This proposal should be either: (1) tempered with strong caveats, or (2) bolstered by additional experimental data. One experiment that potentially could be performed quickly would be to compare burst sizes for infection of host cells lacking type I restriction endonucleases (available commercially) with bacteriophage T7 derivatives having or lacking the gene encoding Ocr; If the authors proposal is true, there should be measurable differences in burst sizes. The reviewers strongly encourage, but do not absolutely require, the authors to attempt to compare burst sizes of ocr+ and ocr= T7 phage in hosts strains lacking type I restriction.

We appreciate the suggested experiments and in fact earlier experiments carried out by McAllister and Barrett, 1977, to some extent have addressed this issue. In that study, the authors used host strains lacking the host restriction enzymes and looked at the effects of host RNA and protein synthesis post T7 infection. Their results showed that expression of gp0.3 (Ocr) resulted in an inhibition, albeit delayed, of host β-galactosidase synthesis and delayed inhibition of total RNA synthesis (Figure 3 in McAllister and Barrett, 1977). This inhibitory effect is consistent with our results here showing that over-expressed Ocr inhibits host transcription, and thus also protein production. The delayed response could be due to the fact that Ocr is expressed by the host RNAP, it thus takes time for sufficient amount of Ocr to be synthesized by the host to show an inhibitory response. We have added this in the text. Interestingly, when Ocr and gp1.1 (a small protein with unknown function) were both expressed in T7, the host protein synthesis was shutdown effectively, suggesting that gp1.1 and Ocr might have complementary/synergetic effects although that requires further investigation. We have added this in Discussion but have also tuned down on the proposal.

Figure Revisions:

Figure 1:

1) In A, there are two peaks of the RNAP-Ocr complex eluting between 40-57 ml. Do the authors know what the difference is between the complexes in both peaks? Which peak was used for structural studies?

Both peaks contain RNAP-Ocr as indicated by SDS-PAGE. They could reflect the two distinct conformations we observe. The peaks were pulled together.

2) In C, it is unclear why the RNAP-Ocr complex is loaded twice in the western blot. What's the difference between both lanes?

They are fractions from the two peaks, again showing both containing Ocr. We have marked this and made clear in figure legend.

3) In D-E, the authors use EC50 and Kd interchangeably. To avoid confusion, the authors should be consistent and refer to Kd. Moreover, considering the labeled ligand in these experiments is at 25 nM concentration and the Kd for the interactions is in a similar concentration range, the binding measurements are performed under conditions of RNAP depletion (i.e. the free concentration of RNAP does not equal the concentration of added RNAP for each data point). If the authors cannot use ligand concentration 10-fold below the Kd for technical reasons, they should use a binding model that accounts for this change in free RNAP concentration or at least acknowledge that the reported Kds are higher than the true Kds.

We appreciate and agree with the reviewers on these points. Indeed, the signal from the labelled protein is a limiting factor on the concentrations we could use in these experiments. We have now refitted the data using the full tight binding equation. A tight binding equation was used with σ or Ocr fixed at 25 nM. The Kd values now are Ocr-RNAP: 7 nM, σ⁷⁰-RNAP: 15 nM and σ⁵⁴-RNAP: 20 nM. These numbers are broadly in agreement with Kd of σ⁵⁴-RNAP (~ 60 nM) obtained in our laboratory (Scott et al. Biochem J., 2000, 352:539-547) using kinetic binding experiments. The competition experiments in Figure 6 (and Figure 6—figure supplement 1) show that equimolar concentration of Ocr can effectively compete out σ⁷⁰ and σ⁵⁴, again consistent with similar Kds to RNAP. Nevertheless, we have made the description more qualitative in the text and made it clear that the Kd values represent maximum values.

4) The authors also state that Ocr is much more abundant in T7 infected cells than type I restriction enzymes, arguing that the additional Ocr molecules interact with PNAP. It would be helpful if the authors could comment on the intracellular concentration of Ocr compared to σ factors and whether these are in a range where Ocr can effectively compete with σ factors for binding to RNAP.

It is difficult to pin down exact numbers but RNAP is estimated to be ~1000 and is in excess of σ⁷⁰ (~ 500) and σ⁵⁴ (< 100) (see https://pax-db.org/dataset/511145/137/). Based on experiments conducted in one of the authors (DFTD)’s lab in terms of SDS-PAGE gel intensities, there are at least a few hundreds of Ocr molecules, compared to < 60 restriction enzymes in infected cells.

Figure 2:

5) Should be revised to compare Ocr-RNAP binding with σ⁷⁰-holoenzyme transcription initiation complexes.

We have done so.

6) For non-experts, it is not easily evident where the template and non-template DNA strands, as well as the RNA, are located in the structure. The authors should consider coloring these elements differently (also in Figure 3).

We have now colored them differently.

Figure 2—figure supplement 1:

7) The authors used focused 3D classification but do not explain how the masks were generated. A description should be included in the Materials and methods section, and it would be useful to also show the masks used in the figure.

We have now added the masks in the figures and detailed how the mask is generated.

Figure 2—figure supplement 2:

8) A and B, the particle orientation distributions are far from isotropic, the effect of this on the FSC curves should be assessed with directional FSC plots (https://3dfsc.salk.edu/). Some of the density maps shown in Figure 2—figure supplement 3 may suffer from this.

Yes the dataset does suffer from preferential orientation issues and the directional FSC plots confirm that. We thank the reviewers for recommendations. We have conducted the analysis and included these plots in Figure 2—figure supplement 2. Furthermore, we have discussed these issues in the text.

9) In A and B, the display threshold for rendering the EM volumes seems too low. Increasing the threshold will likely help to bring out the secondary structure features that should be easily identifiable at this resolution but are not well visible in the representations shown in the figures.

We have replaced the figures.

10) In D, what's the difference between both half-map FSCs? Presumably one map was used for model refinement and the other as control, but it is not clear which is which.

Yes the reviewers are correct and we have now clearly labelled and explained in figure legends.

Figure 2—figure supplement 3:

11) The authors state that many side chain densities are resolved in the map, but this is not well visible in panel C.

We have replaced with new figures to better reflect the quality in the core region.

Figure 4:

12) Should be revised to compare Ocr-RNAP binding with σ⁷⁰-holoenzyme transcription initiation complexes.

We have done so.

13) The authors state in the text (subsection “Structures of Ocr in complex with RNAP”, last paragraph) that the downstream DNA binding channel provides a positively charged surface composed of b, b' clamps and b' jaw, but this is not shown in the figure. The authors further state that positively charged areas on the Ocr surface face the b' jaw domain, but this is also not evident in this figure as the b' jaw is not shown in B. It would be helpful to revise the figure accordingly.

We have made new figures (Figure 4) to illustrate these features.

Figure 7 and Figure 7—figure supplement 1:

14) How were the gels stained in the different panels?

Some were stained for proteins using Coomassie blue, some were western blots (anti-his tag) and visualized by chemical colorimetric exposing. Some were visualized by detecting the Cy3 flourophore signals covalently linked to DNA. This information is now provided in Figure Legends and/or Materials and methods.

15) In Figure 7A and B, the authors argue that there is a gradual disappearance of the RNAP-σ⁵⁴ complex when adding increasing concentrations of Ocr. However, in conditions I and II in 7A there seems to be no or very little RNAP-σ⁵⁴ complex present at either Ocr concentration. A more careful titration would be necessary to make this point. The low abundance of RNAP-σ⁵⁴ in lanes 7 and 9 is also surprising since one would expect a similar abundance of both RNAP-Ocr and RNAP-σ⁵⁴ at equimolar concentration of the components given the similar Kds for the interactions. In condition III, the authors add Ocr to preformed RNAP-σ⁵⁴ purified by gel filtration, but it seems there is barely any σ⁵⁴ present (see also the very weak signal in Figure 7—figure supplement 1A when comparing lanes 11 and 12 to lane 6).

The Kd for Ocr is actually slightly better than for both σ⁷⁰ and σ⁵⁴ (Figure 1D-F). Furthermore, σ does not stain so well. They are more visible in western blotted gels against His-tag on σ proteins (new Figure 6B). We have also moved σ⁵⁴ data to supplementary figures (new Figure 6—figure supplement 1).

Therefore, it is unclear whether Ocr actually competes with σ⁵⁴ or binds to abundant free RNAP. To substantiate their claim, the authors should demonstrate that the preformed RNAP-σ complexes contain stoichiometric amounts of all components as expected. More generally, the gels shown could also benefit from quantitation.

We have shown that the holoenzyme is formed stoichiometrically (Figure 1—figure supplement 1). We have quantified gels when possible.

16) In 7C and 7D, the authors changed the protein concentrations for the different reactions compared to A and B. To ease comparison with the other panels in this figure, it would make more sense to use similar experimental conditions as in the previous experiments. It would also be helpful if the authors could comment on the affinity of the DNA for RNAP-σ. Is binding tighter than that of Ocr to RNAP to explain that DNA can overcome the inhibitory effect of Ocr that is pre-bound to RNAP?

We have conducted experiments under the same conditions as in vitro transcription assays and stained for Cy3 fluorescently labelled DNA (new Figure 6C-D compared to new Figure 5). Figure 6A-B shows competition experiments and protein concentrations are 2x as in Figure 6C-D due to different detection methods. In Figure 6A, the gels were stained with Coomassie blue. However σ factors were not sensitive to Coomassie. Figure 6B shows western blots against His-tag on σ. Figure 6C-D investigates the effects on promoter complex formation, similar to those in Figure 5. Here we used fluorescent signals from Cy3 attached to DNA, which are more sensitive than Coomassie and hence lower protein concentrations were used. We now also include data showing DNA binding to RNAP-σ⁷⁰ and indeed the Kd is ~ 0.1 nM, significantly tighter binding compared to Ocr. These data are included in Figure 6—figure supplement 1.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Figure 6 in the revised manuscript is problematic – the discussion of Figure 6 in the text, in the Figure 6 legend, and the figure itself are inconsistent. While Figure 6 contains panels A – D, the figure legend only refers to panels A – C. In addition, the legend does not adequately describe the figures. The legend to panel A should describe the experiment more completely and explain what Roman numerals I, II and III denote in that panel, as well as in panel B. It would perhaps be best to relabel the panels to denote the reaction scheme as panel A and to explicitly point the reader to that scheme in explaining I, II and III. The differences between the pairs of lanes 7-8, 9-10, 11-12 should also be explained.

We have now amended Figure 6 legend as well as the text (subsection “Ocr inhibits RNAP recruitment and open complex formation but does not 271 disrupt pre-formed open complexes”) in the revised manuscript. I hope now the manuscript can be accepted for publication in eLife.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Figure 6 and the legend are now improved in that one can better understand what is in the figure based on the legend. In panels C and D the authors helpfully added labels above the gel lanes to explain the differences between the pairs of lanes in 'I', 'II', and 'III' (0.5 vs. 5 microM); however, the same needs to be done for panels A and B. Please label panels A and B with the concentrations of Ocr above the pairs of lanes in I, II, and III so that all the panels in the figure are labeled optimally and the same.

We have now amended Figure 6 as well as Figure 6—figure supplement 1 with Ocr concentrations clearly labelled above the pairs of lanes.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Citations

Ye FZ, Zhang XD. 2020. Structural basis of transcription inhibition by the DNA mimic Ocr protein of bacteriophage T7. RCSB Protein Data Bank. 6R9G [DOI] [PMC free article] [PubMed]
Ye FZ, Zhang XD. 2020. Structural basis of transcription inhibition by the DNA mimic Ocr protein of bacteriophage T7. RCSB Protein Data Bank. 6R9B [DOI] [PMC free article] [PubMed]

Supplementary Materials

Supplementary file 1. Pairwise comparisons of E. coli cells expressing Ocr (WT/pBAD18cm[ocr]) versus cells not expressing Ocr (WT/pBAD18cm).

elife-52125-supp1.docx^{(14.4KB, docx)}

Transparent reporting form

elife-52125-transrepform.docx^{(245.8KB, docx)}

Data Availability Statement

The following datasets were generated:

Ye FZ, Zhang XD. 2020. Structural basis of transcription inhibition by the DNA mimic Ocr protein of bacteriophage T7. RCSB Protein Data Bank. 6R9G

Ye FZ, Zhang XD. 2020. Structural basis of transcription inhibition by the DNA mimic Ocr protein of bacteriophage T7. RCSB Protein Data Bank. 6R9B

[bib1] Afonine PV, Grosse-Kunstleve RW, Echols N, Headd JJ, Moriarty NW, Mustyakimov M, Terwilliger TC, Urzhumtsev A, Zwart PH, Adams PD. Towards automated crystallographic structure refinement with phenix.refine. Acta Crystallographica Section D Biological Crystallography. 2012;68:352–367. doi: 10.1107/S0907444912001308. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib2] Atanasiu C, Su TJ, Sturrock SS, Dryden DT. Interaction of the ocr gene 0.3 protein of bacteriophage T7 with EcoKI restriction/modification enzyme. Nucleic Acids Research. 2002;30:3936–3944. doi: 10.1093/nar/gkf518. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib3] Bae B, Davis E, Brown D, Campbell EA, Wigneshweraraj S, Darst SA. Phage T7 Gp2 inhibition of Escherichia coli RNA polymerase involves misappropriation of σ70 domain 1.1. PNAS. 2013;110:19772–19777. doi: 10.1073/pnas.1314576110. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib4] Bae B, Feklistov A, Lass-Napiorkowska A, Landick R, Darst SA. Structure of a bacterial RNA polymerase holoenzyme open promoter complex. eLife. 2015;4:e08504. doi: 10.7554/eLife.08504. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib5] Bandyopadhyay PK, Studier FW, Hamilton DL, Yuan R. Inhibition of the type I restriction-modification enzymes EcoB and EcoK by the gene 0.3 protein of bacteriophage T7. Journal of Molecular Biology. 1985;182:567–578. doi: 10.1016/0022-2836(85)90242-6. [DOI] [PubMed] [Google Scholar]

[bib6] Bennett SE, Schimerlik MI, Mosbaugh DW. Kinetics of the uracil-DNA glycosylase/inhibitor protein association ung interaction with ugi, nucleic acids, and uracil compounds. The Journal of Biological Chemistry. 1993;268:26879–26885. [PubMed] [Google Scholar]

[bib7] Browning DF, Busby SJ. The regulation of bacterial transcription initiation. Nature Reviews Microbiology. 2004;2:57–65. doi: 10.1038/nrmicro787. [DOI] [PubMed] [Google Scholar]

[bib8] Browning DF, Busby SJ. Local and global regulation of transcription initiation in Bacteria. Nature Reviews Microbiology. 2016;14:638–650. doi: 10.1038/nrmicro.2016.103. [DOI] [PubMed] [Google Scholar]

[bib9] Burrows PC, Joly N, Buck M. A prehydrolysis state of an AAA+ ATPase supports transcription activation of an enhancer-dependent RNA polymerase. PNAS. 2010;107:9376–9381. doi: 10.1073/pnas.1001188107. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib10] Cámara B, Liu M, Reynolds J, Shadrin A, Liu B, Kwok K, Simpson P, Weinzierl R, Severinov K, Cota E, Matthews S, Wigneshweraraj SR. T7 phage protein Gp2 inhibits the Escherichia coli RNA polymerase by antagonizing stable DNA strand separation near the transcription start site. PNAS. 2010;107:2247–2252. doi: 10.1073/pnas.0907908107. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib11] Chakraborty A, Wang D, Ebright YW, Korlann Y, Kortkhonjia E, Kim T, Chowdhury S, Wigneshweraraj S, Irschik H, Jansen R, Nixon BT, Knight J, Weiss S, Ebright RH. Opening and closing of the bacterial RNA polymerase clamp. Science. 2012;337:591–595. doi: 10.1126/science.1218716. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib12] Chaney M, Buck M. The sigma 54 DNA-binding domain includes a determinant of enhancer responsiveness. Molecular Microbiology. 1999;33:1200–1209. doi: 10.1046/j.1365-2958.1999.01566.x. [DOI] [PubMed] [Google Scholar]

[bib13] Decker KB, Hinton DM. Transcription regulation at the core: similarities among bacterial, archaeal, and eukaryotic RNA polymerases. Annual Review of Microbiology. 2013;67:113–139. doi: 10.1146/annurev-micro-092412-155756. [DOI] [PubMed] [Google Scholar]

[bib14] Dryden DT. DNA mimicry by proteins and the control of enzymatic activity on DNA. Trends in Biotechnology. 2006;24:378–382. doi: 10.1016/j.tibtech.2006.06.004. [DOI] [PubMed] [Google Scholar]

[bib15] Emsley P, Lohkamp B, Scott WG, Cowtan K. Features and development of coot. Acta Crystallographica. Section D, Biological Crystallography. 2010;66:486–501. doi: 10.1107/S0907444910007493. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib16] Feklístov A, Sharon BD, Darst SA, Gross CA. Bacterial sigma factors: a historical, structural, and genomic perspective. Annual Review of Microbiology. 2014;68:357–376. doi: 10.1146/annurev-micro-092412-155737. [DOI] [PubMed] [Google Scholar]

[bib17] Glyde R, Ye F, Jovanovic M, Kotta-Loizou I, Buck M, Zhang X. Structures of bacterial RNA polymerase complexes reveal the mechanism of DNA loading and transcription initiation. Molecular Cell. 2018;70:1111–1120. doi: 10.1016/j.molcel.2018.05.021. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib18] Goddard TD, Huang CC, Ferrin TE. Visualizing density maps with UCSF chimera. Journal of Structural Biology. 2007;157:281–287. doi: 10.1016/j.jsb.2006.06.010. [DOI] [PubMed] [Google Scholar]

[bib19] Hahn S, Young ET. Transcriptional regulation in Saccharomyces cerevisiae: transcription factor regulation and function, mechanisms of initiation, and roles of activators and coactivators. Genetics. 2011;189:705–736. doi: 10.1534/genetics.111.127019. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib20] Hausmann R, Messerschmid M. Inhibition of gene expression of T7-related phages by prophage P1. Molecular and General Genetics MGG. 1988;212:543–547. doi: 10.1007/BF00330862. [DOI] [PubMed] [Google Scholar]

[bib21] Ho MX, Hudson BP, Das K, Arnold E, Ebright RH. Structures of RNA polymerase–antibiotic complexes. Current Opinion in Structural Biology. 2009;19:715–723. doi: 10.1016/j.sbi.2009.10.010. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib22] Ho CH, Wang HC, Ko TP, Chang YC, Wang AH. The T4 phage DNA mimic protein arn inhibits the DNA binding activity of the bacterial Histone-like protein H-NS. The Journal of Biological Chemistry. 2014;289:27046–27054. doi: 10.1074/jbc.M114.590851. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib23] Issinger OG, Hausmann R. Synthesis of phage coded products during infection of E. coli in gene-1-mutants of phages T3 and T7. Hoppe-Seyler's Zeitschrift Fur Physiologische Chemie. 1972;353:1530. [PubMed] [Google Scholar]

[bib24] Kelleher JE, Raleigh EA. Response to UV damage by four Escherichia coli K-12 restriction systems. Journal of Bacteriology. 1994;176:5888–5896. doi: 10.1128/JB.176.19.5888-5896.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib25] Kennaway CK, Obarska-Kosinska A, White JH, Tuszynska I, Cooper LP, Bujnicki JM, Trinick J, Dryden DT. The structure of m.ecoki type I DNA methyltransferase with a DNA mimic antirestriction protein. Nucleic Acids Research. 2009;37:762–770. doi: 10.1093/nar/gkn988. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib26] Kennaway CK, Taylor JE, Song CF, Potrzebowski W, Nicholson W, White JH, Swiderska A, Obarska-Kosinska A, Callow P, Cooper LP, Roberts GA, Artero JB, Bujnicki JM, Trinick J, Kneale GG, Dryden DT. Structure and operation of the DNA-translocating type I DNA restriction enzymes. Genes & Development. 2012;26:92–104. doi: 10.1101/gad.179085.111. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib27] Kornberg RD. Mechanism and regulation of yeast RNA polymerase II transcription. Cold Spring Harbor. 1998;63:229–232. doi: 10.1101/sqb.1998.63.229. [DOI] [PubMed] [Google Scholar]

[bib28] Krüger DH, Schroeder C. Bacteriophage T3 and bacteriophage T7 virus-host cell interactions. Microbiological Reviews. 1981;45:9–51. doi: 10.1128/MMBR.45.1.9-51.1981. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib29] Ma C, Yang X, Lewis PJ. Bacterial transcription as a target for antibacterial drug development. Microbiology and Molecular Biology Reviews. 2016;80:139–160. doi: 10.1128/MMBR.00055-15. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib30] McAllister WT, Barrett CL. Roles of the early genes of bacteriophage T7 in shutoff of host macromolecular synthesis. Journal of Virology. 1977;23:543–553. doi: 10.1128/JVI.23.3.543-553.1977. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib31] Murshudov GN, Skubák P, Lebedev AA, Pannu NS, Steiner RA, Nicholls RA, Winn MD, Long F, Vagin AA. REFMAC5 for the refinement of macromolecular crystal structures. Acta Crystallographica Section D Biological Crystallography. 2011;67:355–367. doi: 10.1107/S0907444911001314. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib32] Nechaev S, Severinov K. Inhibition of Escherichia coli RNA polymerase by bacteriophage T7 gene 2 protein. Journal of Molecular Biology. 1999;289:815–826. doi: 10.1006/jmbi.1999.2782. [DOI] [PubMed] [Google Scholar]

[bib33] Ooi WY, Murayama Y, Mekler V, Minakhin L, Severinov K, Yokoyama S, Sekine SI. A Thermus phage protein inhibits host RNA polymerase by preventing template DNA strand loading during open promoter complex formation. Nucleic Acids Research. 2018;46:431–441. doi: 10.1093/nar/gkx1162. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib34] Punjani A, Rubinstein JL, Fleet DJ, Brubaker MA. cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nature Methods. 2017;14:290–296. doi: 10.1038/nmeth.4169. [DOI] [PubMed] [Google Scholar]

[bib35] Putnam CD, Tainer JA. Protein mimicry of DNA and pathway regulation. DNA Repair. 2005;4:1410–1420. doi: 10.1016/j.dnarep.2005.08.007. [DOI] [PubMed] [Google Scholar]

[bib36] Ratner D. The interaction bacterial and phage proteins with immobilized Escherichia coli RNA polymerase. Journal of Molecular Biology. 1974;88:373–383. doi: 10.1016/0022-2836(74)90488-4. [DOI] [PubMed] [Google Scholar]

[bib37] Roberts GA, Stephanou AS, Kanwar N, Dawson A, Cooper LP, Chen K, Nutley M, Cooper A, Blakely GW, Dryden DT. Exploring the DNA mimicry of the ocr protein of phage T7. Nucleic Acids Research. 2012;40:8129–8143. doi: 10.1093/nar/gks516. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib38] Scheres SH. RELION: implementation of a bayesian approach to cryo-EM structure determination. Journal of Structural Biology. 2012;180:519–530. doi: 10.1016/j.jsb.2012.09.006. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib39] Srivastava A, Talaue M, Liu S, Degen D, Ebright RY, Sineva E, Chakraborty A, Druzhinin SY, Chatterjee S, Mukhopadhyay J, Ebright YW, Zozula A, Shen J, Sengupta S, Niedfeldt RR, Xin C, Kaneko T, Irschik H, Jansen R, Donadio S, Connell N, Ebright RH. New target for inhibition of bacterial RNA polymerase: 'switch region'. Current Opinion in Microbiology. 2011;14:532–543. doi: 10.1016/j.mib.2011.07.030. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib40] Studier FW. Bacteriophage T7. Science. 1972;176:367–376. doi: 10.1126/science.176.4033.367. [DOI] [PubMed] [Google Scholar]

[bib41] Studier FW. Gene 0.3 of bacteriophage T7 acts to overcome the DNA restriction system of the host. Journal of Molecular Biology. 1975;94:283–295. doi: 10.1016/0022-2836(75)90083-2. [DOI] [PubMed] [Google Scholar]

[bib42] Sturrock SS, Dryden DT, Atanasiu C, Dornan J, Bruce S, Cronshaw A, Taylor P, Walkinshaw MD. Crystallization and preliminary X-ray analysis of ocr, the product of gene 0.3 of bacteriophage T7. Acta Crystallographica Section D Biological Crystallography. 2001;57:1652–1654. doi: 10.1107/S0907444901011623. [DOI] [PubMed] [Google Scholar]

[bib43] Tabib-Salazar A, Liu B, Shadrin A, Burchell L, Wang Z, Wang Z, Goren MG, Yosef I, Qimron U, Severinov K, Matthews SJ, Wigneshweraraj S. Full shut-off of Escherichia coli RNA-polymerase by T7 phage requires a small phage-encoded DNA-binding protein. Nucleic Acids Research. 2017;45:7697–7707. doi: 10.1093/nar/gkx370. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib44] Tabib-Salazar A, Liu B, Barker D, Burchell L, Qimron U, Matthews SJ, Wigneshweraraj S. T7 phage factor required for managing RpoS in Escherichia coli. PNAS. 2018;115:E5353–E5362. doi: 10.1073/pnas.1800429115. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib45] Tagami S, Sekine S, Minakhin L, Esyunina D, Akasaka R, Shirouzu M, Kulbachinskiy A, Severinov K, Yokoyama S. Structural basis for promoter specificity switching of RNA polymerase by a phage factor. Genes & Development. 2014;28:521–531. doi: 10.1101/gad.233916.113. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib46] Tan YZ, Baldwin PR, Davis JH, Williamson JR, Potter CS, Carragher B, Lyumkis D. Addressing preferred specimen orientation in single-particle cryo-EM through tilting. Nature Methods. 2017;14:793–796. doi: 10.1038/nmeth.4347. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib47] Trabuco LG, Villa E, Schreiner E, Harrison CB, Schulten K. Molecular dynamics flexible fitting: a practical guide to combine cryo-electron microscopy and X-ray crystallography. Methods. 2009;49:174–180. doi: 10.1016/j.ymeth.2009.04.005. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib48] Vassylyev DG, Vassylyeva MN, Perederina A, Tahirov TH, Artsimovitch I. Structural basis for transcription elongation by bacterial RNA polymerase. Nature. 2007;448:157–162. doi: 10.1038/nature05932. [DOI] [PubMed] [Google Scholar]

[bib49] Walkinshaw MD, Taylor P, Sturrock SS, Atanasiu C, Berge T, Henderson RM, Edwardson JM, Dryden DT. Structure of ocr from bacteriophage T7, a protein that mimics B-form DNA. Molecular Cell. 2002;9:187–194. doi: 10.1016/S1097-2765(02)00435-5. [DOI] [PubMed] [Google Scholar]

[bib50] Wang HC, Chou CC, Hsu KC, Lee CH, Wang AH. New paradigm of functional regulation by DNA mimic proteins: recent updates. IUBMB Life. 2018;72:539–548. doi: 10.1002/iub.1992. [DOI] [PubMed] [Google Scholar]

[bib51] Wilkinson M, Troman L, Wan Nur Ismah WA, Chaban Y, Avison MB, Dillingham MS, Wigley DB. Structural basis for the inhibition of RecBCD by gam and its synergistic antibacterial effect with quinolones. eLife. 2016;5:e22963. doi: 10.7554/eLife.22963. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib52] Yang Y, Darbari VC, Zhang N, Lu D, Glyde R, Wang YP, Winkelman JT, Gourse RL, Murakami KS, Buck M, Zhang X. Transcription structures of the rna polymerase-σ54 reveal new and conserved regulatory strategies. Science. 2015;349:882–885. doi: 10.1126/science.aab1478. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib53] Zhang Y, Feng Y, Chatterjee S, Tuske S, Ho MX, Arnold E, Ebright RH. Structural basis of transcription initiation. Science. 2012;338:1076–1080. doi: 10.1126/science.1227786. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib54] Zhang K. Gctf: real-time CTF determination and correction. Journal of Structural Biology. 2016;193:1–12. doi: 10.1016/j.jsb.2015.11.003. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib55] Zheng SQ, Palovcak E, Armache J-P, Verba KA, Cheng Y, Agard DA. MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nature Methods. 2017;14:331–332. doi: 10.1038/nmeth.4193. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib56] Zuo Y, Steitz TA. Crystal structures of the E. coli transcription initiation complexes with a complete bubble. Molecular Cell. 2015;58:534–540. doi: 10.1016/j.molcel.2015.03.010. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Structural basis of transcription inhibition by the DNA mimic protein Ocr of bacteriophage T7

Fuzhou Ye

Ioly Kotta-Loizou

Milija Jovanovic

Xiaojiao Liu

David TF Dryden

Martin Buck

Xiaodong Zhang

Roles

Abstract

eLife digest

Introduction

Results

Ocr interacts with RNAP with high affinity

Figure 1. RNAP interacts with Ocr and can form a stable complex with Ocr.

Figure 1—figure supplement 1. Protein purification and complex formation.

Structures of Ocr in complex with RNAP

Figure 2. Structures of RNAP-Ocr in two different binding modes in two orthogonal views.

Figure 2—figure supplement 1. cryoEM data quality and Image processing flowchart.

Figure 2—figure supplement 2. Quality of the two distinct RNAP-Ocr complex 3D reconstructions.

Figure 2—figure supplement 3. Representative electron density maps in the two reconstructions.

Figure 2—figure supplement 4. Structural comparison of Ocr from RNAP-Ocr complex (magenta and yellow) with that from crystal structure (blue, PDB code 1S7Z).

Figure 3. Detailed interactions between Ocr and RNAP and in the ‘wide-clamp’ conformation.

Figure 4. Detailed interactions between RNAP and Ocr in the ‘narrow-clamp’ conformation.

Ocr inhibits transcription in vitro and in cells

Figure 5. In vitro transcription and in vivo reporter assays.

Figure 5—figure supplement 1. In vitro transcription assays of σ54-dependent systems using linear (A) and supercoiled DNA (B) and σ70 transcription using super-coiled DNA (C).

Figure 5—figure supplement 2. in vivo data showing Ocr can inhibit transcription.

Ocr inhibits RNAP recruitment and open complex formation but does not disrupt pre-formed open complexes

Figure 6. Ocr and its effect on σ70 holoenzyme and open complex formation.

Figure 6—figure supplement 1. Ocr interferes with holoenzyme formation of σ54(A-B).

Discussion

Ocr is a bifunctional protein

Figure 7. Comparisons with sigma and other phage proteins in binding to RNAP.

Comparisons with other RNAP inhibiting bacteriophage proteins

Unique features of Ocr defines its ability in specific DNA processing systems

Materials and methods

RNAP-Ocr complex formation

Microscale thermophoresis (MST)

CryoEM sample preparation

Electron microscopy data collection

Image processing

Model building, refinement and structural analysis

Table 1. Cryo-EM data collection and refinement statistics.

Native gel mobility assay and western blotting

Small primed RNA assays

Bacterial growth curves

Beta-galactosidase assays

RNA extraction, DNase digestion, reverse transcription and quantitative polymerase chain reaction (RT-qPCR)

Acknowledgements

Funding Statement

Contributor Information

Funding Information

Additional information

Competing interests

Author contributions

Additional files

Data availability

References

Decision letter

Roles

Author response

Associated Data

Data Citations

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Figure 5—figure supplement 1. In vitro transcription assays of σ⁵⁴-dependent systems using linear (A) and supercoiled DNA (B) and σ⁷⁰ transcription using super-coiled DNA (C).

Figure 6. Ocr and its effect on σ⁷⁰ holoenzyme and open complex formation.

Figure 6—figure supplement 1. Ocr interferes with holoenzyme formation of σ⁵⁴(A-B).