Structural basis for transcription inhibition by E. coli SspA

Abstract

Stringent starvation protein A (SspA) is an RNA polymerase (RNAP)-associated protein involved in nucleotide metabolism, acid tolerance and virulence of bacteria. Despite extensive biochemical and genetic analyses, the precise regulatory role of SspA in transcription is still unknown, in part, because of a lack of structural information for bacterial RNAP in complex with SspA. Here, we report a 3.68 Å cryo-EM structure of an Escherichia coli RNAP-promoter open complex (RPo) with SspA. Unexpectedly, the structure reveals that SspA binds to the E. coli σ⁷⁰-RNAP holoenzyme as a homodimer, interacting with σ⁷⁰ region 4 and the zinc binding domain of EcoRNAP β′ subunit simultaneously. Results from fluorescent polarization assays indicate the specific interactions between SspA and σ⁷⁰ region 4 confer its σ selectivity, thereby avoiding its interactions with σ^s or other alternative σ factors. In addition, results from in vitro transcription assays verify that SspA inhibits transcription probably through suppressing promoter escape. Together, the results here provide a foundation for understanding the unique physiological function of SspA in transcription regulation in bacteria.

INTRODUCTION

Bacterial transcription initiation is carried out by a bacterial RNA polymerase (RNAP) holoenzyme comprising the bacterial RNAP core enzyme (subunit composition α₂ββ′ω) and a σ factor (1). The primary σ factor (group-1 σ factor; σ⁷⁰ in Escherichia coli; σ^A in Gram-positive bacteria) mediates transcription initiation at most housekeeping genes required for growth under normal conditions, while alternative σ factors, such as σ^s, direct the transcription of stress genes in response to metabolic, developmental and environmental signals. In addition, various transcription regulatory proteins also associate with RNAP in a dynamic manner to modulate the activity of RNAP in the process of transcription initiation, transcription elongation, and transcription termination (1).

The E. coli protein SspA was identified as a bacterial RNA polymerase-associated protein about 40 years ago and its expression was induced by glucose, nitrogen, phosphate or amino acid starvation (2,3). The reported physiological functions of E. coli SspA include responding to changes in the NTP pool of the cell through regulating the NTP kinase activities (4) and establishing the stationary phase-induced acid tolerance by downregulating the cellular level of H-NS (5–7). SspA is highly conserved among Gram-negative bacteria. Its orthologues in Neisseria gonorrhoeae, Francisella novicida, Francisella tularensis and Vibrio cholerae were shown to affect the expression of genes involved in pathogenesis (8–15). Intriguingly, SspA-like proteins are also present in higher organisms. Significant homologies between SspA and several stress- or auxin-regulated plant proteins have been also reported, suggesting that SspA may be a member of a highly conserved group of stress-induced proteins (3).

The crystal structure of SspA from Haemophilus influenza, Pseudomonas fluorescens, Pseudomonas putida and Yersinia pestis have been determined so far (16). SspA belongs to the cytosolic glutathione transferase (GST) family based on its structural similarity to canonical GST proteins, which are usually composed of two domains–a thioredoxin-like N-terminal domain and a larger C-terminal domain. The N-terminal domain constitutes the majority of GSH binding site, while the C-terminal domain contains the binding pocket for hydrophobic co-substrates (16). Although SspA is structurally similar to canonical GST proteins, it lacks the glutathione transferase activity and is different in the oligomerization. Instead, SspA regulates transcription in various bacteria by directly contacting RNAP through a conserved ‘PHP’ motif (17). Although many functional, biochemical, and structural studies of SspA have been performed in recent four decades, the molecular mechanism and structural basis of SspA-mediated transcription regulation is still unknown, partially because of a lack of precise structural information for bacterial RNAP in complex with SspA (3,4,6–10,12,17–20).

In this report, to gain insight into the functional role of SspA in transcription, we determined a single-particle cryo-electron microscopy (cryo-EM) structure of a E. coli RNAP-promoter open complex (RPo) with SspA at 3.68 Å resolution. The structure reveals that SspA bridges RNAP core enzyme and σ⁷⁰ by making interactions both with zinc binding domain (ZBD) of RNAP-β′ subunit and with a non-conserved patch on region 4 of the primary σ⁷⁰ factor (σ⁷⁰R4). Further biochemical results confirmed the interactions between SspA and E. coli σ⁷⁰-RNAP and show that SspA inhibits transcription activity of E. coli σ⁷⁰-RNAP. Our structure with the biochemical results suggest SspA as a global transcription repressor of σ⁷⁰ and such inhibition is probably through suppressing promoter escape. The results here provide the structural basis of SspA–RNAP interaction and a foundation for understanding the unique physiological function of SspA in transcription regulation of bacteria.

MATERIALS AND METHODS

SspA protein

Gene encoding E. coli sspA was synthesized and subcloned to pET28a by Sangon Biotech, Inc. E. coli strain BL21(DE3) (Invitrogen, Inc.) was transformed with plasmid pET28a-NH-SspA (Sangon Biotech, Inc) encoding N hexahistidine-tagged SspA under the control of the bacteriophage T7 promoter. Single colonies of the resulting transformants were used to inoculate 1 l LB broth containing 50 μg/ml kanamycin, cultures were incubated at 37°C with shaking until OD₆₀₀ = 0.6, cultures were induced by addition of isopropyl-β-d-thiogalactoside (IPTG) to 0.5 mM, and cultures were incubated at 20°C overnight. Then, cells were harvested by centrifugation (5000 rpm; 10 min at 4°C), resuspended in 20 ml buffer A (20 mM Tris–HCl, pH 8.0, 0.1 M NaCl,5% glycerol) and lysed using a ATS AH-10013 cell disrupter (ATS, Inc.). The lysate was centrifuged (12 000 rpm; 45 min at 4°C), and the supernatant was loaded onto a 2 ml column of Ni-NTA agarose (Qiagen, Inc.) equilibrated with buffer A. The column was washed with 10 ml buffer A containing 0.16 M imidazole and eluted with 10 ml buffer A containing 0.5 M imidazole. The sample was further purified by anion-exchange chromatography on a Mono Q 10/100 GL column (GE Healthcare, Inc.; 160 ml linear gradient of 0.1–1.0 M NaCl in buffer A). Fractions containing SspA were pooled and stored at −80°C. SspA derivatives were expressed and purified in the same way as wild-type protein. The protein concentration is determined by using a BCA protein assay kit (Pierce™ BCA Protein Assay Kit, Thermo Scientific™, Inc.). Yields were ∼2 mg/l, and purities were >95%.

Alanine-substituted SspA derivatives were prepared as described for preparation of SspA, but using plasmid pET28a-SspA derivatives constructed using site-directed mutagenesis (QuikChange Site-Directed Mutagenesis Kit; Agilent).

E. coli σ⁷⁰ protein

Escherichia coli σ⁷⁰ protein was prepared using plasmid pGEMD as reported (21). Yield was ∼50 mg/l, and purity was >95%.

E. coli RNAP core enzyme

Escherichia coli RNAP core enzyme was prepared from E. coli strain BL21(DE3) (Invitrogen, Inc.) transformed with plasmid pIA900, using culture, induction and purification procedures essentially as reported (21). Yield was ∼2.5 mg/l, and purity was >95%.

Assembly of E. coli SspA–RPo complex

DNA oligonucleotides (sequences in Figure 1A) (Sangon Biotech, Inc.) were dissolved in nuclease-free water to ∼1 mM and stored at −80°C. Template strand DNA and nontemplate strand DNA were annealed at a 1:1 ratio in 10 mM Tris–HCl, pH 7.9, 0.2 M NaCl and stored at −80°C. E. coli SspA-RPo was prepared in reaction mixtures containing (500 μl): 9 μM σ⁷⁰, 18 μM SspA, 4.5 μM E. coli RNAP core enzyme and 5 μM DNA scaffold. E. coli σ⁷⁰ protein was incubated with SspA for 10 min at 37°C, incubated with core for 10 min at 37°C and incubated with DNA scaffold for 10 min at 37°C. The mixture was applied to a Superose 6 Increase 10/300 GL column (GE Healthcare, Inc.) equilibrated in 10 mM HEPES, pH 7.5, 50 mM KCl, and the column was eluted with 24 ml of the same buffer. Fractions were checked by SDS-PAGE and the peak containing E. coli SspA-RPo complex was concentrated to 20 μM using an Amicon Ultra-0.5 ml centrifugal filter (10 kDa MWCO; Merck Millipore, Inc.).

Figure 1.

The overall structure of E. coli RNAP-promoter open complex with SspA. (A) The scaffold used in structure determination of E. coli RNAP-promoter open complex with SspA (SspA–RPo). (B, C) The orthogonal view orientations of the cryo-EM density map (B) and structure model (C) of E. coli transcription initiation complex with SspA. The RNAP, SspA and nucleic acids are presented as cartoon and colored as indicated in the color key. The density map is shown in gray envelop. The cryo-EM density map (blue transparent surface) for SspA. The cryo-EM density (red and yellow transparent surface) for the scaffold DNA and σ⁷⁰. NT, non-template-strand promoter DNA; T, template-strand promoter DNA.

Cryo-EM grid preparation

Immediately before freezing, 8 mM CHAPSO was added to the sample. C-flat grids (CF-1.2/1.3-4C; Protochips, Inc.) were glow-discharged for 60 s at 15 mA prior to the application of 3 μl of the complex, then plunge-frozen in liquid ethane using a Vitrobot (FEI, Inc.) with 95% chamber humidity at 10°C.

Cryo-EM data acquisition and processing

The grids were imaged using a 300 kV Titan Krios (FEI, Inc.) equipped with a K2 Summit direct electron detector (Gatan, Inc.). Images were recorded with Serial EM in counting mode with a physical pixel size of 1.307 Å and a defocus range of 1.5–2.5 μm. Data were collected with a dose of 8 e/pixel/s. Images were recorded with a 12 s exposure and 0.25 s subframes to give a total dose of 59 e/Ų. Subframes were aligned and summed using MotionCor2 (22). The contrast transfer function was estimated for each summed image using CTFFIND4 (23). From the summed images, ∼10 000 particles were manually picked and subjected to 2D classification in RELION (24). 2D averages of the best classes were used as templates for auto-picking in RELION. Auto-picked particles were manually inspected, then subjected to 2D classification in RELION. Poorly populated classes were removed, resulting in a dataset of 610 066 particles. These particles were 3D classified in RELION using a map of E. coli RNAP-promoter open complex low-pass filtered to 40 Å resolution as a reference (25). 3D classification resulted in four classes. Particles in Class 1 (RPo) were 3D auto-refined and post-processed in RELION. Particles in Class 3 were subjected to an additional 3D classification and the class with density for SspA was 3D auto-refined and post-processed.

Cryo-EM model building and refinement

The models of RNAP core enzyme, σ⁷⁰ and DNA scaffold from the structure of E. coli RPo (PDB 6CA0) (25), the crystal structure of SspA (PDB 1YY7) were fitted into the cryo-EM density map using Chimera (17,26). The model of nucleic acids was built manually in Coot (27). The coordinates were real-space refined with secondary structure restraints in Phenix (28).

Fluorescence polarization assay

The procedures of this fluorescence polarization assay were followed as previous reported with essential modifications (29–31). The reaction mixtures (100 μl) contain the fluorescein-labeled wild type SspA or SspA mutant derivatives (100 nM; final concentration) in FP buffer (PBS buffer) were incubated for 10 min at room temperature. E. coli σ⁷⁰-RNAP holoenzyme or E. coli σ^S-RNAP holoenzyme (a serial of final concentrations including 0, 3.5, 7, 14, 28, 56, 72, 96, 112, 120, 144, 168, 192, 216 and 240 μM) was added and incubated for 10 min at room temperature. The FP signals were measured using a plate reader (SPARK, TECAN, Inc.) equipped with an excitation filter of 485/20 nm and an emission filter of 520/20 nm. The data were plotted in SigmaPlot14.0 (Systat software, Inc) and the dissociation constant K_d was estimated by fitting the data to the following equation:

where F is the FP signal at a given concentration of RNAP, F₀ is the FP signal in the absence of RNAP, [S] is the concentration of RNAP and B is an unconstrained constant, Error bars represent mean± SEM out of n = 3 experiments.

In vitro multi-round transcription assays

Multi-round runoff transcription assays were carried out using E. coli σ⁷⁰-RNAPholoenzyme and wild-type EcoSspA or mutant derivatives through a mango method. For wild-type SspA concentration-dependent transcription activity measurements, the reaction (40 μl) was performed at 37°C and contained 100 nM N25 promoter Mango DNA, 100 nM E. coli σ⁷⁰-RNAP holoenzyme in reaction buffer [50 mM Tris–HCl, pH 8.0, 10 mM MgCl₂, 0.5% (vol/vol) glycerol, 100 mM potassium chloride, 1 mM DTT, 0.1% Tween-20]. Reaction was pre-incubated at 37°C for 5 min, and then 10 μl EcoSspA (a series of final concentrations including 0, 1.6, 3.2, 6.4, 12.8, 25.6 and 51.2 μM) was added before the addition of the NTP mix (0.1 mM; final concentration) and Tol-biotin (0.5 mM; final concentration) incubated for 30 min. The fluorescence signals were measured using a plate reader (SPARK, TECAN, Inc.) at an excitation wavelength of 510 nm and an emission wavelength of 550 nm. The data were plotted in SigmaPlot14.0 (Systat software, Inc.), Error bars represent mean± SEM out of n = 3 experiments. For evaluating the relative transcription activities between wild-type SspA and its mutants, the procedures and reaction system were similar as wild-type SspA concentration-dependent transcription activity measurements shown above except that wild-type SspA and its mutants concentration (25.6 μM final concentration) were added into the reaction system.

Electrophoretic mobility shift assay (EMSA) of E. coli RNAP-promoter open complex(RPo) with SspA

Template strand DNA oligonucleotide (5′-TCCCCTGCATCCGTGACAGCTCCCATTATAGC ACAATTTAACACTTTTGTCAATCATTTTGTT-3′, Sangon Biotech, Inc.) and non-template strand DNA oligonucleotide (5′- AACAAAATGATTGACAAAAGTGTTAAATTGTGCTAT AATGGGAGCTGTCACGGATGCAGGGGA-3′, Sangon Biotech, Inc.) were annealed at a 1:1 ratio in 10 mM Tris–HCl, pH 7.9, 0.2 M NaCl and stored at −80°C. Electrophoretic mobility shift assays were performed in reaction mixtures containing (20 μl): 0.4 μM wild type SspA or SspA mutant derivatives, 0.1 μM E. coli σ⁷⁰-RNAPholoenzyme, 0.05 μM DNA scaffold, 0.1mg/ml heparin, 7 mM Tris–HCl (pH 7.9), 50 mM Tris–Ac (pH 7.9), 0.19 M KGlu, 5 mM MgAc₂, 0.4 mM EDTA,0.2 mM DTT, 0.125 mg/ml BSA, 50 mM potassium phosphate (pH 6.5), 0.14 M NaCl and 22% glycerol. Wild type SspA protein incubated with E. coli σ⁷⁰-RNAP holoenzyme for 10 min at 37°C, then incubated with DNA scaffold for 10 min at 37°C, and incubated with 0.1 mg/ml heparin for 1 min at 37°C. The reaction mixtures were applied to 5% polyacrylamide slab gels (29:1 acrylamide/bisacrylamide), electrophoresed in 90 mM Tris-borate, pH 8.0, and 0.2 mM EDTA, stained with 4S Red Plus Nucleic Acid Stain (Sangon Biotech, Inc.) according to the procedure of the manufacturer, and analyzed by ImageJ (https://imagej.nih.gov/ij/).

Data analysis

Data for fluorescence polarization assays and in vitro multi-rounds transcription assays are means of three technical replicates. Error bars represent mean± SEM out of n = 3 experiments.

RESULTS

Overall structure of E. coli RNAP-promoter open complex (RPo) with SspA

The Cryo-EM structure of E. coli RNAP-promoter open complex with SspA (SspA-RPo) was determined using a recombinant E. coli RNAPholoenzyme, a recombinant E. coli SspA, and a synthetic nucleic-acid scaffold comprising a 34-bp upstream dsDNA, a 16-bp downstream dsDNA, and a noncomplementary transcription bubble with a consensus -10 element (Figure 1A) (21,32,33). The SDS-PAGE result confirmed that all protein components are present in the complex (Supplementary Figure S1B); and the EMSA result confirmed that the protein complex was able to bind the synthetic nucleic-acid scaffold DNA (Supplementary Figure S1C). The cryo-EM dataset was collected on Titan Krios and the particles were classified into 4 classes after 3D classification. The third Class, which shows clear presence of SspA and thereby represents SspA-RPo, was subjected 3D auto-refinement/post-process and finally refined to a 3.68 Å nominal resolution (Table 1; Figure 1B and C; Supplementary Figures S2–S5). The first class, which represents a regular RPo, was refined to a 3.58 Å nominal resolution (Table S1; Supplementary Figure S6 and S7). The cryo-EM map of SspA-RPo shows clear signals for E. coli σ⁷⁰RNAP holoenzyme, SspA and nucleic-acid scaffold (Supplementary Figure S5). The structures of EcoRNAP core enzyme and σ⁷⁰ from SspA-RPo structure are very similar to the previously reported E. coli RNAP-promoter open complex (E. coli RPo) structure with a root-mean-square deviation (RMSD) of 1.33 Å (3694 Cαs aligned) (25). The region 2 of σ⁷⁰ (σ⁷⁰R2) and the region 4 of σ⁷⁰ (σ⁷⁰R4) interact with the conserved promoter –10 element and –35 element in a similar manner as they do in E. coli RPo. The crystal structure of Yersinia pestis SspA could be readily fit into the map, suggesting little conformational change of SspA upon interaction with E. coli σ⁷⁰-RNAPholoenzyme (Figure 1B and Supplementary Figure S2A).

Table 1.

Single particle cryo-EM data collection, processing and model building for Escherichia coli RNAP-promoter open complex (RPo) with SspA

Data collection and processing
Microscope	Titan Krios
Voltage (kV)	300
Detector	K2 summit
Electron exposure (e/Å2)	59
Defocus range (μm)	1.5–2.5
Data collection mode	counting
Physical pixel size (Å/pixel)	1.307
Symmetry imposed	C1
Initial particle images	614,193
Final particle images	60,145
Map resolution (Å)a	3.68
Refinement
Map sharpening B-factor (Å)	–104
Root-mean-square deviation
Bond length (Å)	0.007
Bond angle (°)	0.782
Molprobity statistics	2.93
Clashscore	11.70
Rotamer outliers (%)	10.60
Cβ outliers (%)	0.0
Ramachandran plot	99.88
Favored (%)	87.38
Outliers (%)	0.12

^aGold-standard FSC 0.143 cutoff criteria.

The Cryo-EM structure of SspA–RPo clearly shows that E. coli SspA homodimer locates on the surface of the E. coli σ⁷⁰-RNAPholoenzyme (Figure 1B, C; Supplementary Figure S2). One SspA protomer (SspA I) mainly interacts with σ⁷⁰R4 through a large interface of ∼483.5 Ų and the other SspA protomer (SspA II) contacts ZBD from RNAP-β′ subunit through an interface of ∼320.5 Ų (Figure 1B) (34). Both SspA I and SspA II approach the upstream edge of –35 element but make no direct contact with the promoter DNA (Figures 1B, C). Such interaction mode between SspA and σ⁷⁰-RPo supports the previous finding that SspA can bind to either RNAP core enzyme or RNAP holoenzyme using biochemical and genetic approaches (2,3,7,17). In addition, the Cryo-EM structure of SspA–RPo clearly reveals the interface of the SspA homodimer, which mainly involves the helices α4−α5 from both protomers and buries about 30% of the total surface area. The interface in our SspA-RPo structure is similar to that of Y. pestis SspA alone (Supplementary Figures S8A) (17), hinting that such homodimer interface is conserved and essential for physiological activities of SspA. Residues Arg73, Glu77, Tyr78, Glu81, Arg96 and Arg100 from α4 and α5 of both protomers form salt bridge and hydrogen bond networks stabilizing the interface. These hydrophilic residues are further surrounded by hydrophobic residues Leu69, Val94, Gly97 and Leu101, which protrude from α4 and α5 helices of SspA I and SspA II, strengthening the homodimer interactions (Supplementary Figures S8B). All of these structural observations are consistent with previous biochemical and genetic results (17), and match our fluorescence polarization assay result showing that G97I mutation disrupts the binding of E. coli SspA to RNAP holoenzyme (Supplementary Figures S8C) (12).

Interactions between SspA and σ⁷⁰R4

In the cryo-EM structure of E. coli SspA-RPo, neither of SspA protomers contacts the non-template or template DNA, indicating that SspA may not affect the RNAP activities through interacting with the promoter directly as other typical transcription factors, such as CRP or NtrC of E. coli (35–38). The interactions between SspA I and σ⁷⁰R4 include a polar interaction network, which consists of Arg65, Arg82, Pro84, His85 from SspA I and Asn568, Lys578, Gln579, Asp584 from σ⁷⁰R4, and van del Waals interactions between residues of SspA I (Phe83, Pro84, His 85 and Pro86) and a hydrophobic shallow groove composed of several σ⁷⁰R4 residues (Met561, Asn568, Lys557, Met567 and Tyr571 (Figure 2C). Specifically, residue His85, one of the highly conserved ‘PHP’ motif residues in SspA (17), is embedded into the shallow hydrophobic groove on σ⁷⁰R4 (Figure 2A–C). Notably, most evolutionarily conserved residues of SspA, especially the ‘PHP’ motif, are clustered in the interface, implicating a functional relevance of its interaction with σ⁷⁰(Figure 2C). Moreover, the E. coli SspA derivative P84A/H85A/P86A loses the ability to support the acid resistance of E. coli, further suggesting the physiological importance of the SspA/σ⁷⁰R4 interface (17).

Figure 2.

The interactions between SspA and the σ⁷⁰R4. (A) The relative location of SspA, σ⁷⁰R4 and upstream double strand DNA in the structure of E. coli RNAP-promoter open complex with SspA. (B) SspA interacted with negative surface of σ⁷⁰. The electrostatic potential surface of σ⁷⁰ R4 was generated using APBS tools in Pymol. SspA is represented as a blue cartoon. (C) The detailed interactions between the σ⁷⁰ R4 and SspA(stereo view). Hydrogen bonds are shown as red dashed lines.

Interactions between SspA and the ZBD of RNAP-β′ subunit

The structure also reveals specific interactions between SspA and the ZBD of RNAP-β′ subunit (Figure 3A). These interactions include, 1) direct polar interaction network made by Ser35, Arg65, Arg82 and His85 from SspA II and Ser32, Arg60, Glu86 and Glu91 from ZBD domain of RNAP-β′ subunit; 2) van der Waals interactions made by Phe83 and Pro84 from SspA II, Tyr 92 and Pro93 from SspA I, and Ser34, Arg81, Val83 and Ile84 from ZBD domain of RNAP-β′ subunit (Figure 3B).

Figure 3.

The interactions between SspA and the ZBD of RNAP-β′ subunit. (A) The relative location among SspA, σ⁷⁰R4 and upstream double strand DNA. SspA II interacted with the zinc binding domain (ZBD) from RNAP-β′ subunit. SspA II is represented as a blue transparent surface and cartoon; ZBD from RNAP-β′ subunit is represented as cartoon. (B) The detailed interactions between the β′ subunit and SspA (stereo view). Salt-bridge bonds are shown as red dashed lines.

We subsequently evaluated contribution of the interface residues to the SspA-σ⁷⁰ and the SspA–ZBD interactions using fluorescence polarization assay. Previous studies suggested that a conserved ‘PHP’ motif within SspA is critical for the function of SspA and the binding of SspA to RNAP (17). In our structure, the ‘PHP’ motif plays an indispensable role in the interaction between SspA and σ⁷⁰, constituting part of the interface. Our results further show the triple mutation P84A/H85A/P86A of the ‘PHP’ motif, as well as single mutations (R65A, R82A, P84A, H85A or P86A), significantly impaired the SspA–RNAP interaction, validating our structure and highlighting the significance of the SspA-σ⁷⁰ interface (Figure 4A, B). Our fluorescence polarization assay assessing effects of all possible alanine substitutions at the SspA-ZBD interface, including R65A, R82A, P84A, H85A and Y92A, also confirmed the significance of these residues observed in the cryo-EM structure (Figure 4B). Intriguingly, most evolutionarily conserved residues of SspA are clustered in the interfaces of SspA–ZBD and SspA-σ⁷⁰R4, implicating a functional relevance of the interaction of SspA with σ⁷⁰ and RNAP-β′ subunit (Figure 4C).

Figure 4.

The interactions between SspA, the σ⁷⁰R4 and the ZBD of RNAP-β′ subunit: binding affinity data. (A) Binding affinities between wild-type SspA or SspA-H85A and σ⁷⁰-RNAP measured by a fluorescence polarization (FP) assay. varying amounts of the σ⁷⁰-RNAP as indicated (mean ± SEM; three determinations). (B) Relative binding affinities of wild-type SspA and its mutants from the σ⁷⁰R4-SspA interface or ZBD of β′ subunit-SspA interface measured by the fluorescence polarization assay (mean ± SEM; three determinations). Error bars represent mean± SEM out of n = 3 experiments. (C) Protein Sequence Alignments of SspA from ∼100 non-redundant bacterial species. The sequences were extracted from UniProt Database by BLAST. The alignment was performed by Cluster Omega and the sequence logos were generated on the WebLogo server (http://weblogo.berkeley.edu/logo.cgi). Black filled circles, residues involved in interactions with the ZBD from RNAP-β′ subunit; Red filled circles indicate residues that are involved with interactions with σ⁷⁰R4. The residues are numbered as in E. coli SspA.

SspA inhibits transcription by stabilizing the association of σ⁷⁰R4 with RNAP core enzyme

A previous report showed that Pseudomonas aeruginosa SspA could function as an anti-σ⁷⁰ factor involving in transcription regulation of Alginate production (20). In our cryo-EM structure of SspA-RPo, E. coli SspA interacts with σ⁷⁰R4 and ZBD simultaneously (Figure 5A), suggesting SspA may act as a bridge and enhance the interactions between σ⁷⁰R4 and ZBD of RNAP-β′ subunit. Considering that σ⁷⁰R4 serves as the hub for docking class II transcription activators but has to dissociate during promoter escape, we next explored the transcription output of the SspA–RNAP interaction.

Figure 5.

SspA inhibits promoter escape by interacting with the σ⁷⁰R4 and β′ subunit of RNAP simultaneously. (A) Overall structure of E. coli σ⁷⁰-RNAP-promoter open complex shown in surface. SspA dimer was shown in cartoon. For clarity, α, β, β' and ω were represented as gray surface, other colors as in Figure 1. (B) Wild-type SspA concentration-dependent transcription activities evaluated by in vitro multi-rounds transcription assay. Each 40 μl reaction contains the E. coli σ⁷⁰-RNAP holoenzyme (100 nM), the fluorescence labelled mango N25 promoter DNA template (100 nM), and varying amounts of the wild-type SspA as indicated (mean ± SEM; three determinations). (C) Relative transcription activities of wild-type SspA and its mutants from the interface between SspA and ZBD of β′ subunit to E. coli RNAP holoenzyme evaluated by the in vitro multi-rounds transcription assay (mean ± SEM; three determinations). Error bars represent mean± SEM out of n = 3 experiments.

To explore the effect of SspA on RNAP activities, we modified an fluorescence-based in vitro multi-round transcription assay (30,39). The results in Figure 5B show clear concentration-dependent transcription inhibition by the wild-type SspA, suggesting that SspA behaves as a transcription repressor. We subsequently evaluated the effect on transcription by alanine-substitution derivatives of SspA, which showed defect on binding to σ⁷⁰R4 or ZBD (Figures 4B). The results showed that most of the tested mutations (R65A, F83A, P84A, P86A, Y92A and P84A/H85A/P86A) reduced the transcription inhibition ability, suggesting that SspA-RNAP interaction accounts for the transcription inhibition of SspA (Figure 5C). The results lead to a hypothesis that SspA glues σ⁷⁰R4 and RNAP core enzyme, and inhibits promoter escape, a process requiring dissociation of both promoter and σ⁷⁰R4 from RNAP core enzyme (1,39–43)

SspA specifically inhibits transcription from σ⁷⁰-RNAP holoenzyme

Having demonstrated that SspA functions as a transcription repressor to inhibit σ⁷⁰-dependent gene transcription. We next asked whether SspA interacts with other alternative σ factors in E. coli and inhibits transcription initiated by these alternative σ factors. Sequence alignment of all six σ factors of E. coli reveals that the surface corresponding to SspA-interacting patch on σ⁷⁰R4 is not conserved among σ⁷⁰ and other alternative σ factors; even for σ^S, the master stress σ factor that is most closely related σ⁷⁰ in sequence and structure, there are six key interface residues (Lys557 versus Glu272, Met561 vs. Arg276, Met567 versus Gly282, Asn568 versus Tyr283, His571 versus Ala286, Asp581 versus Gly296; σ⁷⁰ versus σ^S) different from those of σ⁷⁰R4 (Figure 6A–C). The sequence comparison suggests that SspA probably is a σ⁷⁰-specific transcription repressor. To validate such hypothesis, we tested the interaction between SspA and σ^s-RNAP holoenzyme. The results from fluorescence polarization assays clearly showed that SspA interacts with the σ^s-RNAP holoenzyme with a much lower affinity than that of σ⁷⁰-RNAP holoenzyme (Figure 6D), supporting SspA as a σ⁷⁰-specific transcription repressor. Furthermore, we also mutated the σ⁷⁰ residue Lys557, Met567, Asn568, Tyr571, Gln579 and Asp581 to alanine in order to evaluate their effects on the binding affinities of SspA and σ⁷⁰-RNAP holoenzyme. The results clearly exhibited that these residues are critical for the SspA and σ⁷⁰ interactions (Figure 6E), which is consistent with the conclusion that SspA may act as a σ⁷⁰-specific transcription repressor.

Figure 6.

Protein–protein interactions between σ⁷⁰ and σ^S and SspA. (A) Structure-based sequence alignment of potential SspA-interacting residues in E. coli σ⁷⁰ and other alternative σ factors of E. coli. The non-conserved interaction residues are indicated with magenta stars; The secondary structure elements of σ⁷⁰ is shown at the top. (B) Protein–protein interaction between σ⁷⁰R4 and SspA. (C) Protein–protein interaction between σ^SR4 and SspA. SspA, σ⁷⁰R4 and σ^SR4 are represented as cartoon colored in blue, yellow and wheat, respectively. (D) Binding affinities between σ⁷⁰-RNAP or σ^S-RNAP and wild-type SspA measured by a fluorescence polarization assay. varying amounts of the σ⁷⁰-RNAP or σ^S-RNAP as indicated (mean ± SEM; three determinations). (E) Relative binding affinities of wild-type σ⁷⁰ and its region 4 mutants to E.coli RNAP core enzyme and wild-type SspA measured by the fluorescence polarization assay. Error bars represent mean± SEM out of n = 3 experiments.

DISCUSSION

SspA was discovered as an RNAP-associated protein ∼40 years ago. However, the physiological function of SspA in gene expression remains largely unclear. A large collection of biochemical, biophysical and genetic data imply that SspA may regulate transcription in an unprecedented manner (2,3,6–12,17–20). SspA is associated with the virulence of several pathogenic bacteria including F. tularensis, N. gonorrhoeae, V. cholerae and enterohaemorrhagic E. coli (EHEC) (6,8,9,15). E. coli SspA was shown to be required for acid resistance and transcriptional activation of phage P1 late genes. A recent report also suggested that P. aeruginosa SspA may function as an anti-σ⁷⁰ factor (20). In this study, we show that SspA inhibits transcription activity of RNAP-σ⁷⁰ holoenzyme and provide structural explanations for such inhibition.

Our cryo-EM structure of E. coli SspA-RPo shows that SspA acts as a stabilizing chaperon connecting σ⁷⁰₄ and RNAP–β′ ZBD of E. coli RNAP core enzyme but does not contact promoter DNA (Figure 1). The interface between SspA and RNAP-σ⁷⁰ holoenzyme is relatively large (∼483.5 Ų in total) and comprises both hydrophobic and polar interactions. The interaction mode of RNAP-σ⁷⁰ holoenzyme and SspA is in sharp contrast to the interaction mode of RNAP-σ⁷⁰ holoenzyme and canonical class II transcription activators, which typically make interaction with small activation patches on σ⁷⁰R4 through electrostatic interactions. We infer such difference of the interaction mode of canonical transcription activators and SspA to RNAP accounts for the difference of the consequences of their transcription regulation. The weak electrostatic interactions allow DNA-bound transcription activators to efficiently dock on RNAP-σ⁷⁰ holoenzyme at the stage of RPo formation and to dissociate from the RNAP-σ⁷⁰ holoenzyme at the stage of prompter escape without much obstacle. However, the large interface, made by both hydrophobic and hydrophilic interactions, between SspA and RNAP-σ⁷⁰ holoenzyme tends to glue the σ⁷⁰R4 and RNAP core enzyme together, restricts the flexibility of σ⁷⁰R4 domain, which is believed to undergo substantial conformational change resulting in the dissociation of σ⁷⁰R4 during promoter escape process (1,39–43).

Most bacterial transcription factors repress gene expression by binding to DNA targets that overlap essential elements at their target promoters, thereby occluding access of RNAP (44). In many cases, repression is enhanced by the binding of multiple transcription repressor molecules at some promoters, which bind distally but interact with each other via DNA loops (44–46). At other promoters subjected to repression, RNAP is able to engage but is blocked at the promoter by the transcription repressor (44–46). A few transcription repressors, such as the CytR repressor, are anti-activators that simultaneously interact with their operator and adjacent activators, such as the cyclic AMP (cAMP) receptor protein. At some promoters, CytR binding requires a combination of CytR–CRP and CytR–DNA interactions to prevent the binding of RNA polymerase (35,38). Therefore, it is widely accepted that transcription factors repress transcription mainly by sterically occluding the transcription machinery on promoter DNA (44,46).

However, there are also few examples of transcription repression requiring directly interaction between a repressor and RNA polymerase. For example, the P4 protein encoded by phage φ29, which infects B. subtilis, simultaneously binds to the C-terminal domain of the α-subunit of RNA polymerase and to the DNA upstream of the polymerase, thereby preventing promoter clearance (47). The SspA in principal fits into this category of transcription repressors, but differs in protein fold, RNAP-interacting mode, and probably the sub-steps of promoter escape that SspA acts on.

Collectively, we revealed here that SspA decreases the transcriptional activity of σ⁷⁰-RNAP in a DNA contact-independent manner and through stabilizing the key structural elements of σ⁷⁰ and RNAP-β’subunit. Our study provides the structural basis and molecular mechanism of an unprecedented example of transcription repression. The transcription effect of SspA—repressing σ⁷⁰-dependent gene expression and facilitating σ^s-dependent stress-related gene expression—would help σ^s-RNAP to outcompete the transcription activity of housekeeping σ⁷⁰-RNAP and substantially increase the transcription activity of the σ^s-RNAP holoenzyme in the expression of stress-related genes (Figure 7). The unique DNA contact-independent mechanism also provides a new paradigm for bacterial transcription repression.

Figure 7.

Proposed working models of SspA. SspA inhibits the promoter escape process via interaction with σ⁷⁰R4 but not inhibit the promoter escape due to the absence of interaction with σ^SR4 under stress conditions.

DATA AVAILABILITY

Atomic coordinates and structure factors for the cryo-EM structures of E. coli σ⁷⁰ RNAP holoenzyme-SspA RNAP-promoter open complex has been deposited into the PDB and EMDB with accession codes PDB 7C97 and EMDB 30307, respectively. Atomic coordinates and structure factors for the cryo-EM structures of E. coli RNAP-promoter open complex has been deposited into the PDB and EMDB with accession codes PDB 7CHW and EMDB 30376, respectively.

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online.

FUNDING

National Natural Science Foundation of China [81903526, 81991523 to W.L., in part); Fok Ying Tung Education Foundation, Natural Science Foundation of Jiangsu Province of China [BK20190798 to W.L.]; The Open Project of State Key Laboratory of Natural Medicines [SKLNMKF202004 to W.L.]; Jiangsu Specially-Appointed Professor Talent Program (to W.L.). Funding for open access charge: Natural Science Foundation of China [81903526, 81991523 to W.L.]; Fok Ying Tung Education Foundation, Natural Science Foundation of Jiangsu Province of China [BK20190798 to W.L.].

Conflict of interest statement. None declared.