Architectural principles for Hfq/Crc-mediated regulation of gene expression

Xue Yuan Pei; Tom Dendooven; Elisabeth Sonnleitner; Shaoxia Chen; Udo Bläsi; Ben F Luisi

doi:10.7554/eLife.43158

. 2019 Feb 13;8:e43158. doi: 10.7554/eLife.43158

Architectural principles for Hfq/Crc-mediated regulation of gene expression

Xue Yuan Pei ^1,^†, Tom Dendooven ^1,^†, Elisabeth Sonnleitner ², Shaoxia Chen ³, Udo Bläsi ^2,^✉, Ben F Luisi ^1,^✉

Editors: Anna Marie Pyle⁴, John Kuriyan⁵

PMCID: PMC6422490 PMID: 30758287

Abstract

In diverse bacterial species, the global regulator Hfq contributes to post-transcriptional networks that control expression of numerous genes. Hfq of the opportunistic pathogen Pseudomonas aeruginosa inhibits translation of target transcripts by forming a regulatory complex with the catabolite repression protein Crc. This repressive complex acts as part of an intricate mechanism of preferred nutrient utilisation. We describe high-resolution cryo-EM structures of the assembly of Hfq and Crc bound to the translation initiation site of a target mRNA. The core of the assembly is formed through interactions of two cognate RNAs, two Hfq hexamers and a Crc pair. Additional Crc protomers are recruited to the core to generate higher-order assemblies with demonstrated regulatory activity in vivo. This study reveals how Hfq cooperates with a partner protein to regulate translation, and provides a structural basis for an RNA code that guides global regulators to interact cooperatively and regulate different RNA targets.

Research organism: Other

eLife digest

Living things can adapt rapidly to changes in their surroundings by switching whole groups of genes on and off. These responses must be controlled carefully, and they are often coordinated by regulatory proteins working together. Within a cell, the coded information in genes is copied to create molecules called mRNAs, which are then translated to produce proteins. Stopping the cell from reading the information in mRNAs is one way of shutting down specific genes.

Pseudomonas aeruginosa is a species of bacteria that can infect humans and can cause cases of sepsis and pneumonia. It changes the activity of its genes in response to its environment. For example, certain genes are only active inside a human host. This allows the bacteria to make the best use of available nutrients and energy. In these cells, two proteins named Hfq and Crc cooperate to silence groups of genes. They do this by stopping the cell from reading specific mRNA molecules, but how they do this is not fully understood.

By using a technique called cryo-electron microscopy, Pei, Dendooven et al. studied Hfq and Crc attached to mRNAs. The results show that two groups of six Hfq molecules and two Crcs attach themselves to two mRNA sections to create a structure that stops an mRNA from being translated into a protein. Since the structure only forms with certain mRNAs, the effect is specific to certain genes. The structure needs both Hfq and Crc to work together, which means it only forms in specific situations – when the affected genes are not needed.

P. aeruginosa is highly antibiotic resistant and new drugs are urgently needed to control infections. Understanding how this disease-causing bacterium controls its genes could lead to new treatments. The mechanisms of gene regulation are also common to many other forms of life so may also aid the wider understanding of how cells adapt to rapidly changing environments.

Introduction

The RNA chaperone Hfq contributes to the control of mRNA translation through different modes of action. In one mode, Hfq acts indirectly by facilitating base-pairing interactions of cognate mRNA targets with small regulatory RNAs (Vogel and Luisi, 2011; Wagner and Romby, 2015). In a second mode, the RNA chaperone can directly repress translation independently of small RNAs, by binding A-rich sequences at or in the vicinity of translation initiation sites (Vecerek et al., 2005; Salvail et al., 2013; Sonnleitner and Bläsi, 2014). As a member of the extensive Lsm/Sm protein family, Hfq shares an ancient structural core that oligomerizes to form toroidal architectures exposing several RNA-binding surfaces. Crystallographic and biophysical data showed that RNA recognition is mediated by distinct interactions with distal, proximal and rim faces of the hexameric ring (Schumacher et al., 2002; Link et al., 2009; Sauer et al., 2012; Horstmann et al., 2012; Panja et al., 2013), as well as revealing the role of the unstructured C-terminal tail in auto-regulating RNA-binding activities (Santiago-Frangos et al., 2016; Santiago-Frangos et al., 2017).

In the opportunistic, Gram-negative pathogen Pseudomonas aeruginosa, Hfq acts as a pleiotropic regulator of metabolism (Sonnleitner and Bläsi, 2014), virulence (Sonnleitner et al., 2003; Fernández et al., 2016; Pusic et al., 2016), quorum sensing (Sonnleitner et al., 2006; Yang et al., 2015) and stress responses (Lu et al., 2016). Many of these roles are likely facilitated through partner molecules, and numerous putative protein interactors of P. aeruginosa Hfq have been identified with functions in transcription, translation and mRNA decay (Van den Bossche et al., 2014). In the case of Hfq from E. coli, the functionally important partners include RNA polymerase, ribosomal protein S1 (Sukhodolets and Garges, 2003), the endoribonuclease RNase E (Ikeda et al., 2011), polyA-polymerase, and the exoribonuclease polynucleotide phosphorylase (Mohanty et al., 2004; Bandyra et al., 2016). Most likely, these complexes are RNA mediated and affect the co-localisation of the machineries of transcription, translation and RNA decay (Worrall et al., 2008; Resch et al., 2010; Vecerek et al., 2010).

One P. aeruginosa protein that was found to co-purify with tagged Hfq is the Catabolite repression control protein, Crc (Van den Bossche et al., 2014; Moreno et al., 2015; Sonnleitner et al., 2018). Crc is involved in carbon catabolite repression (CCR) in Pseudomonas, a process that channels metabolism to use preferred carbon sources (such as succinate) until they are exhausted, whereupon alternative sources are used (Rojo, 2010). In addition to carbon catabolite repression, Hfq and Crc link key metabolic and virulence processes in Pseudomonas species. The two proteins affect biofilm formation, motility (O'Toole et al., 2000, Huang et al., 2012; Zhang et al., 2012; Pusic et al., 2016), biosynthesis of the virulence factor pyocyanin (Sonnleitner et al., 2003; Huang et al., 2012), and antibiotic susceptibility (Linares et al., 2010; Heitzinger, 2016). Recent ChiP-seq studies indicate that Hfq and Crc have an even broader regulatory impact in Pseudomonas and can work in concert to bind many nascent transcripts co-translationally, uncovering a large number of potential regulatory targets (Kambara et al., 2018). Crc is a specialized protein that is found in a limited subset of bacteria including Pseudomonads but excluding Enterobacteriaceae such as E. coli. The Crc protein is a member of the EEP family (Exonuclease/Endonuclease/Phosphatase), but the absence of key catalytic residues in the active site cleft indicate that Crc has lost enzymatic function (Milojevic et al., 2013).

Crc and Hfq act together to translationally repress mRNA target genes. The targets are then subjected to degradation, which might trigger disassembly of the Hfq/Crc/RNA complex (Sonnleitner and Bläsi, 2014). The impact of Hfq/Crc on translational repression is countered by the non-coding RNA CrcZ, which increases in levels when preferred carbon sources are exhausted (Sonnleitner et al., 2009; Sonnleitner and Bläsi, 2014). CrcZ expression is under control of the alternative sigma factor RpoN and the two-component system CbrA/B (Sonnleitner et al., 2009). Although the signal responsible for CbrA/B activation remains unknown, it is thought to be related to the cellular energy status (Valentini et al., 2014).

The mechanism by which Hfq and Crc act in translational repression of target mRNAs involves binding of both proteins to ribosome-binding sequences. In the case of the translationally repressed amiE mRNA, encoding aliphatic amidase, the distal face of Hfq binds to an A-rich segment termed amiE_6ARN that comprises the ribosome binding site (Sonnleitner and Bläsi, 2014). In contrast, Crc has no intrinsic RNA-binding activity (Milojevic et al., 2013) but strengthens binding of A-rich target transcripts to the distal side of Hfq (Sonnleitner et al., 2018). How Hfq binds to A-rich RNAs is structurally understood (Link et al., 2009), but the cooperative role of Crc in this context has not been elucidated. To gain insight into how P. aeruginosa Hfq cooperates with Crc in translational repression of amiE, we determined the structure of the complex they form on the Hfq binding motif of the CCR-controlled amiE mRNA using cryo-electron microscopy (cryoEM). Our analyses revealed that the components form higher order assemblies and explain for the first time how a widely occurring structural motif can support the association of Hfq and RNA into cooperative ribonucleoprotein complexes that have regulatory roles. We observe that the interactions supporting the quaternary structure are required for in vivo translational regulation. These findings expand the paradigm for in vivo action of Hfq through cooperation with the Crc helper protein and RNA to form effector assemblies.

Results

An ensemble of Hfq/Crc/amiE_6ARNRNA assemblies

For cryo-EM structural studies of the Hfq/Crc/RNA complex, purified recombinant Hfq and Crc proteins were mixed with an 18 nucleotide Hfq binding motif from the translation initiation region of the CCR-controlled amiE mRNA (hereafter termed amiE_6ARN). The defined sequence (5'-AAA-AAU-AAC-AAC-AAG-AAG-3') has a binding motif comprised of 6 repeats of an A-R-N pattern preferred by the distal face of Hfq (Sonnleitner and Bläsi, 2014). The 18 nucleotides of amiE_6ARN could be distinguished in the cryoEM maps and will be referred to with numbers 1–18 from the 5´ to the 3´ end. The purified sample of Hfq/Crc/amiE_6ARN, after mild chemical crosslinking, yielded well defined single particles on graphene oxide in thin, vitreous ice. Analysis of the reference free 2D class averages and subsequent 3D classification indicated three principal types of complexes corresponding to different stoichiometries of Hfq (hexamer):Crc:amiE_6ARN with compositions 2:2:2, 2:3:2 and 2:4:2 (Figure 1). These higher order assemblies are in agreement with recently observed SEC-MALS and mass spectrometry results which excluded a simple 1:1:1 assembly (Sonnleitner et al., 2018). The maps for the complexes are estimated to be 3.1 Å, 3.4 Å and 3.2 Å in resolution, respectively, based on gold-standard Fourier shell correlations (Figure 1—figure supplement 1). The distribution of the complexes corresponds to roughly 49%, 29% and 23% for the 2:2:2, 2:3:2 and 2:4:2 complexes (Figure 1). The individual crystal structures of Hfq and Crc dock well into the cryoEM densities (Figure 1 and Figure 1—figure supplement 2), and aside from side chain rotations there are few other structural changes of the components (Figure 1—figure supplement 2). To probe for artefacts introduced by the crosslinking experiments, CryoEM analyses were subsequently performed with samples that had not been treated by crosslinking using grid preparation conditions optimised for the crosslinked specimens. Although limited to lower resolution, data from these specimens revealed that the quaternary structure remained unchanged within the resolution limits of the data for the model being compared (Supplementary file 1A).

Figure 1. — Three main classes of particles were observed after reference-free 2D classification (top), corresponding to Hfq/Crc/*amiE*_6ARN assemblies with compositional stoichiometries of 2:2:2, 2:3:2 and 2:4:2. The *amiE*_6ARN species (red) constitute the main interaction interface between Hfq and Crc, together forming the 2:2:2 core complex observable in all three models (bottom). Cyan: Hfq hexamer, orange purple and pink: Crc monomers, red: *amiE*_6ARN. All cryoEM maps were low-pass filtered to 6 Å for clarity of presentation and the crystal structures were docked in as rigid bodies. Only a subset of the high quality 2D classes are shown in the panel.

Figure 1—figure supplement 1. — Three main classes of particles were observed after reference-free 2D classification (top), corresponding to Hfq/Crc/*amiE*_6ARN assemblies with compositional stoichiometries of 2:2:2, 2:3:2 and 2:4:2. The *amiE*_6ARN species (red) constitute the main interaction interface between Hfq and Crc, together forming the 2:2:2 core complex observable in all three models (bottom). Cyan: Hfq hexamer, orange purple and pink: Crc monomers, red: *amiE*_6ARN. All cryoEM maps were low-pass filtered to 6 Å for clarity of presentation and the crystal structures were docked in as rigid bodies. Only a subset of the high quality 2D classes are shown in the panel.

A specialised and recurring RNA conformation in Hfq-mediated regulation

Link et al. (2009) described the crystal structure of E. coli Hfq bound to a polyriboadenylate 18-mer and observed that the RNA encircled the distal face of the Hfq hexamer via a repetitive tripartite binding scheme (Figure 2—figure supplement 1A). Each base triplet is partially embedded between adjacent Hfq monomers and is mostly surface exposed, folding into a’ crown-like’ conformation. We observe striking similarities with the fold of the authentic amiE_6ARN species on the distal side of the P. aeruginosa Hfq hexamer (Figure 2 and Figure 2—figure supplement 1A). Notably, the cryoEM maps were calculated without any reference to Link et al. (2009). The agreement between the co-crystal structure of the analogous complex and the entirely independently derived cryoEM based model provides thus a strong validation.

Figure 2. — 6 RNA triplets are partially embedded in six binding pockets on the Hfq distal side, forming a weaving, crown-like pattern. The A and R sites are occupied by adenine and a purine, respectively, whereas the RNA entry/exit site has no discriminatory preferences and is therefore referred to as the ‘N’ site. The Cryo-EM density for *amiE*_6ARN is depicted as a grey mesh, with the RNA ‘crown’ modelled in red. Positively charged protuberances (blue) guide the RNA to fold into a star-shaped conformation to maximize the surface interaction between the negatively charged RNA backbone, and the positively charged Hfq surface pattern. An atomic model of the A-R-N occupation pattern. Left panel: Adenosine specificity site. Right panel: Purine specificity site. *amiE* nucleotide carbon atoms are depicted in red, Hfq carbon atoms are in green.

Figure 2—figure supplement 1. — 6 RNA triplets are partially embedded in six binding pockets on the Hfq distal side, forming a weaving, crown-like pattern. The A and R sites are occupied by adenine and a purine, respectively, whereas the RNA entry/exit site has no discriminatory preferences and is therefore referred to as the ‘N’ site. The Cryo-EM density for *amiE*_6ARN is depicted as a grey mesh, with the RNA ‘crown’ modelled in red. Positively charged protuberances (blue) guide the RNA to fold into a star-shaped conformation to maximize the surface interaction between the negatively charged RNA backbone, and the positively charged Hfq surface pattern. An atomic model of the A-R-N occupation pattern. Left panel: Adenosine specificity site. Right panel: Purine specificity site. *amiE* nucleotide carbon atoms are depicted in red, Hfq carbon atoms are in green.

Like its E. coli homologue, Pseudomonas Hfq contains six tripartite binding pockets on the distal side, capable of binding a total of 18 nucleotides. Each of the six RNA triplets of the amiE_6ARN RNA fits into an inter-subunit cleft in Hfq (Figure 2). The specific, star-shaped RNA fold is guided by six positively charged protuberances on the distal face of Hfq, with the phosphate backbone circularly weaving in between these, seemingly to minimise steric hindrance while maximizing surface interactions (Figure 2). As described by Link et al. (2009), each pocket consists of an adenosine specificity site (A), a purine nucleotide specificity site (R), and a presumed RNA entrance/exit site (N) which is non-discriminatory (Figure 2—figure supplement 1A). Hfq thus has a structural preference for (ARN)_n RNA stretches on its distal side, where N is any nucleotide. The adenosine specificity (A) sites are organised identically to the corresponding A sites in E. coli Hfq, forming hydrogen bonds between the peptide backbone and carboxyl-groups of Gln33 and the N6,7 atoms of the adenosine base, and a polar interaction between Gln52 (N_ε) and the N1 atom of the adenosine base (Figure 2, Figure 2—figure supplement 1A). The peptide backbone amide of residue Lys31 interacts with the 5’ phosphate group of adenine. Finally, the adenine base is stacked against the side chain of Leu32 (Figure 2). The interactions with N1, by the carbonyl of Q33, and N7, by the amide of Q33, confer pocket-specificity for A as they are not compatible with a G, which would form a repulsive contact via its O6 and peptide carbonyl (Figure 2). The purine (R) specificity site is defined by two neighbouring monomers, where the side chains from Tyr25 and from Leu26’, Ile30’ and Leu32’ (where the prime denotes residues from a neighbouring subunit) contact the nucleotide aromatic base. In amiE_6ARN, one R-site is populated by a guanine, forming a hydrogen bond between the N_ε of Gln52’ and the guanine exocyclic O6 (Figure 2). Just like in the E. coli Hfq/polyA₁₈ structure (Link et al., 2009), Gln52’ forms a physical link between the A and R sites. Previous structures were obtained from polyA RNA, whereas the structures presented here were solved with the authentic amiE Hfq recognition site. Interestingly, Thr61 O_γ forms a double hydrogen bond with the N1 and the exocyclic N2 from the guanine base, which was not seen previously (Link et al., 2009) as all R-sites were occupied by adenine residues (Figure 2). As we will describe further below, the N site base interacts with Crc, but without apparent sequence preference. Strikingly, the A-R-N motif is not supported by Hfq of the Gram-positive bacteria Staphylococcus aureus and Bacillus subtilis, which instead use a R-L (purine, linker) motif (Horstmann et al., 2012) (Figure 2—figure supplement 1B). Phe 30 hinders the RNA backbone, and thus the A-site base, from entering the A-site, while stabilising the R-site together with M32 via stacking interactions.

The quaternary structure of the core complex

In the core complex (with 2:2:2 stoichiometric composition of Hfq:Crc:RNA), two Hfq hexamers sandwich the RNA and Crc (Figure 3A), with each Hfq interacting with one amiE_6ARN RNA and two Crc components. The assembly has C2 symmetry, with the molecular twofold axis passing through the interface of the two Crc molecules. The Crc´s in the assembly self-interact in the same way as observed in the crystal structure of the isolated Crc dimer, where the interface is generated through crystallographic symmetry (Wei et al., 2013; Milojevic et al., 2013). As anticipated from mutational analyses (Sonnleitner et al., 2018), the dominating protein/RNA interaction is made by the distal face of Hfq, forming an interface area of roughly 2270 Å².

Figure 3—figure supplement 1. — (A) Atomic model of the 2:2:2 Hfq/Crc/RNA complex. The view is along the C2 molecular symmetry axis which passes through the homodimeric Crc interface. Hfq hexamers flank the Crc dimer and present the *amiE*_6ARN RNA to form two different interfaces with the Crc protomers, which form an anti-parallel dimer. The Crc protomers form strong polar contacts with mainly the backbone phosphate groups and exposed ribose rings (bottom insets). Two small C2 symmetric binding interfaces constitute the Crc dimerisation (top insets). A single short stretch on each Crc monomer binds a Hfq monomer (middle left panel). Dimeric Crc in yellow and orange, *amiE*_6ARN RNA in red, Hfq hexamers in cyan. (B) Translational regulation of an *amiE::lacZ* reporter gene by Crc variants. *P. aeruginosa* strain PAO1Δcrc(pME9655) harboring plasmids pME4510 (vector control), pME4510crc_Flag (Crc_wt) or derivatives thereof encoding the respective mutant proteins was grown to an OD₆₀₀ of 2.0 in BSM medium supplemented with 40 mM succinate and 40 mM acetamide. The β-galactosidase values conferred by the translational *amiE::lacZ* fusion encoded by plasmid pME9655 in the respective strains are indicated. The results represent data from two independent experiments and are shown as mean and range. (C) Table of Crc-*amiE* contacts in the 2:2:2 assembly. BB: Crc contacts with the RNA backbone; Base: Crc contacts with RNA bases.

The two Crc molecules interact with RNA residues exposed on the surface of Hfq, and both Crc molecules contact the Hfq-rim on the distal side (Figure 3A). The Crc protomers form contacts mainly with the backbone phosphate groups and exposed ribose rings of the RNA (Figure 3A and C). Because the Crc forms antiparallel dimers, there are two modes of interaction of the Crc with amiE_6ARN RNA. In one mode of interaction, Arg140 η¹-NH₂ and Arg141 ε-NH and η¹-NH₂ interact with the phosphodiester backbone of amiE_6ARN. Arg140 and Arg196 form a sandwich with the purine-base of the A3 nucleotide at an Entry/Exit ‘N’ site of amiE_6ARN RNA. Interestingly, Arg140 participates in the Crc1-Crc2 dimer interface in the Crc-dimer as solved by crystallography via contacts with E193 (pdb-ID 4jg3, data not shown), but our maps clearly show a rotamer shift of the R140 side chain away from E193, forming contacts with amiE as described above. The Arg196 ε-NH and η¹-NH₂ groups form hydrogen bonds with the U6-amiE_6ARN backbone and the U6 O2 group forms a hydrogen bond with the Met156 amide. In the second mode of interaction, Lys155 ζ-NH₂ makes a hydrogen bond with the OP₂-group of C9 and the ribose hydroxyl group. Additional hydrogen bonds are formed between Trp161 ε¹-NH and Arg162 η^1/2-NH₂ and the phosphate backbone of amiE_6ARN (Figure 3A).

The highly organised interactions in the core complex (Figure 3A) illustrate how the bases of amiE_6ARN as presented by Hfq constitute a molecular interface for the RNA-mediated interactions between Hfq and Crc. Crc forms small contact surfaces to the RNA, to Hfq, and to itself as a homodimer; these small areas work together to give an assembly that is most likely stabilised through chelate cooperativity. Notably, there is a striking absence of any lower order assemblies in the cryo EM micrographs. The 2:2:2 complex is therefore likely to be the minimal complex formed when all components are present and must be constructed in an ‘all or nothing’ manner, somewhat like a binary switch.

Function, origins and validation of subunit cooperativity in the 2:2:2 complex

The dimer interface of the Crc pair is the largest protein-protein interface in the 2:2:2-complex and has a buried area of 766 Å², which typically corresponds to a moderate intermolecular affinity. The key dimerization interface is maintained by salt bridges between Arg229-Arg230 of one Crc monomer and Glu142 of the second Crc monomer, which is further stabilised by pi-stacking of the Phe231-Phe231 rings across the twofold axis (Figure 3A). The phenylalanine residues are in turn stabilised by stacking interactions with Trp255 of the same Crc monomer (not shown). Two additional polar contacts are formed between Arg137 and the Asn184 carbonyl group of two pairs of helices in the Crc dimer, forming a smaller secondary interface (Figure 3A).

To verify selected interactions between the Crc protomers and Crc and RNA, we explored the effects of Crc variants on Hfq/Crc-mediated repression of an amiE::lacZ translational reporter gene encoded by plasmid pME9655 (Sonnleitner et al., 2018). In addition, an in vitro co-immunoprecipitation assay was employed to assess the capacity of Hfq and Crc variants to form a complex in the presence of amiE_6ARN RNA (Sonnleitner et al., 2018). First, we asked whether R140 (Crc_R140) is required for the interaction of the protein with the RNA (Figure 3A, bottom left inset). As shown in Figure 3B, the Crc_R140E mutant was deficient in translational repression of the amiE::lacZ reporter gene, similarly as observed for the crc deletion strain. Moreover, Crc_R140E did not co-immunoprecipitate with Hfq in the presence of amiE_6ARN RNA (Figure 3—figure supplement 1A), indicating that the interaction between Crc_R140 and RNA is pivotal for Hfq/Crc/RNA complex formation.

Next, we focused on the interactions in the E142 and R229/R230 ‘triangle’ (Figure 3A, top left inset) for the Crc-Crc interaction. The single mutant proteins Crc_R229E and Crc_R230E did not affect translational repression of amiE::lacZ, whereas the function of the Crc_E142R variant was diminished (Figure 3B), indicating that E142 can interact with either R229 or R230. The de-repression of amiE:lacZ observed with the Crc_E142R variant was partially compensated by the double mutant proteins Crc_{E142R, R229E} and Crc_{E142R, R230E} (Figure 3B). In addition, the Crc_E142R and Crc_R230E variants were impaired in Hfq/Crc/RNA complex formation as shown by the co-immunoprecipitation assay (Figure 3—figure supplement 1A). Strikingly, the compensatory changes present in the triple mutant protein Crc_{E142R, R229E, R230E} almost fully restored translational repression of the amiE::lacZ reporter gene (Figure 3B). As the respective Crc variant proteins were produced at comparable levels (Figure 3—figure supplement 1B), these mutational studies support the in vivo role for the interactions of the Crc protomers observed in the cryo-EM models. The exchange of Glu142 with an Ala in the triple mutant protein Crc_{E142A, R229E, R230E} restored as well translational repression of the amiE::lacZ reporter gene (Figure 3B). We interpret this as showing that the rescue of the deleterious mutations does not require charge compensation, but removing any potential charge repulsion.

Recruitment of additional Crc units to the core

The quaternary organisation of the 2:2:2 complex forms a core unit that is also present in the 2:3:2 and 2:4:2 complexes. In that common core, the interaction of Crc with RNA leaves approximately half of the accessible surface of the nucleic acid exposed. For the 2:3:2 and 2:4:2 complexes, additional Crc units are recruited through interactions with the exposed portion of the RNA. As such, the C2 symmetry is broken by the third Crc molecule in the 2:3:2 complex (Figure 1). Interestingly, recruitment of a fourth Crc monomer to the complex restores the C2 symmetry, preserving the symmetry axis from the core complex, but with a conformationally different Crc dimer interface between Crc molecules 3 and 4 (Figure 4A). The two additional Crc monomers have small surface-area contacts with the rest of the complex and are likely to be comparatively mobile, which may account for the stronger variation in resolution for the 2:3:2 and 2:4:2 maps compared to the rather rigid 2:2:2 core assembly (Figure 1—figure supplement 1).

Figure 4. — (A) Atomic model of the 2:4:2 Hfq/Crc/RNA complex. The insets show additional Hfq-Crc, Crc-Crc and Crc-RNA interactions not present in the 2:2:2 complex. The Crc3-4 dimer is formed by only one interface, with an R230-R230 interaction at the core, which globally overlaps with the dimer interface of the Crc1-Crc2 dimer (top left inset). Only one of two RNA-binding patches is presented to *amiE*_6ARN in the Crc3-4 dimer, yet exploited more extensively (right inset). A small interface is formed between Crc3-4 and Hfq. Crc dimer in yellow, *amiE*_6ARN RNA in red, Hfq hexamers in cyan, extra Crc dimer in magenta and purple. (B) Translational regulation of an *amiE::lacZ* reporter gene by Crc variants, as described in Figure 3B. The results represent data from two independent experiments and are shown as mean and range. (C) Two distinct dimeric Crc species are observed in the three complexes solved by cryoEM. i: The self-complementary interaction of the 2:2:2 core complex. ii: In the 2:4:2 complex, an alternative dimer is formed, showing a twisted dimer interface and more open configuration, with Arg230 serving as a dynamic hinge (bottom). (D) An electropositive half-channel runs along the dimer interface of the Crc1-2 dimer, and in the context of the Hfq/Crc/RNA assembly it could potentially serve as a conduit for RNA (dotted white arrow; see Figure 5). The A, R, and E RNA interaction sites of Hfq are annotated.

Function, origins and validation of subunit cooperativity in the 2:4:2 complex

The protomer interactions of the 2:2:2 assembly are highly interdependent, and once the core complex is generated it can apparently recruit additional Crc molecules, forming the 2:3:2 and 2:4:2 complexes. In the 2:4:2 complex, a second type of Crc dimer seems to assemble with a smaller buried surface (Figure 4A). Such additional dimers can only form when an intact 2:2:2 core complex is present, as they are not observed in the core complex itself. Notably, the additional Crc dimer is in a more ‘open’ conformation than the dimer in the core (Figure 4C). The same key Crc dimer interface is occupied but seems to serve as a dynamic hinge, whereas the secondary, smaller, dimer interface between the Crc helices is absent to allow the new Crc dimer to adopt an ‘open’ conformation. Arg230 is reorganised by Glu193 in the same protomer to self-interact with the corresponding Arg230 in the partner Crc, rather than with Glu142 (Video 1). Additional hydrogen bonds are formed between Arg233 and Glu193, whereas Arg229 is no longer part of the dimer interface (Figure 4A). Both Arg230 and Glu193 seem to play important roles in providing the structural freedom to form a dynamic hinge (Figure 4C). Glu193 is part of the Crc1-Crc2 interface in the crystal structure but does not participate in this interface when assembled in the 2:2:2 core complex. Instead it aids in coordinating the Crc3-Crc4 interface in the 2:4:2 assembly. Interestingly, Crc3 and Crc4 do not interact with Crc1 or Crc2, and seem to be recruited solely via Hfq/amiE_6ARN. By reorganising the extra Crc molecules 3 and 4 that bind the 2:2:2 core (Figure 4A), the alternative Crc dimer is able to utilise one of two basic patches on its surface when engaging amiE_6ARN without causing steric hindrance to the already bound Crc dimer.

Video 1. A comparison of the Crc-Crc interface in the 2:2:2 Hfq:Crc:RNA complex (the core complex) and the second type of interface formed between the additional Crc components of the 2:4:2 complex.

Download video file^{(2.5MB, mp4)}

DOI: 10.7554/eLife.43158.011

Open in a new tab

The Arg233-Glu193 interaction is unique for the 2:4:2 assembly and was therefore assessed in vivo. Strikingly, the Crc_E193R mutation fully abrogated translational repression of the amiE:lacZ reporter gene (Figure 4B). The model predicts that the deleterious Crc_E193R mutation can be compensated by the substitution of Crc_R230E to re-establish the interaction. This pair does indeed behave as predicted, as the Crc_{E193R, R230E} variant restored translational repression of the amiE::lacZ reporter gene, further confirming the in vivo importance of the 2:4:2 assembly during CCR (Figure 4B). The exchange of Glu193 with Ala in the Crc_{E193A, R230E} mutant protein also restored translational repression of the reporter gene, albeit to a somewhat reduced extent when compared with the Crc_{E193R, R230E} variant (Figure 4B). Again, these results suggest that the rescue of the deleterious mutations does not require charge compensation, but act by removing any potential charge repulsion.

In addition to Crc Arg140 and Arg141, Crc K139 ζ-NH₂ makes a hydrogen bond with the OP₂-group of A12, Arg138 η¹-NH₂ interacts with the ribose hydroxyl group of C9 and K135 ζ-NH₂ forms a hydrogen bond with the A11 OP₂. Finally, the O2 of cytosine C12 engages in a hydrogen bond with the backbone amino group of Arg140. Direct interactions between the reorganised Crc dimer and Hfq are limited to the same Crc β-strand and exposed loop of a sole Hfq monomer, as in the core complex. Due to the open conformation of the alternative Crc dimer, the Hfq Thr49 hydroxyl group now forms a hydrogen bond with the Ala 78 amide group (Figure 4A).

Interestingly, a basic half-channel is formed over the core dimer interface, with additional basic patches spread over the RNA binding surface of the Crc dimer (Figure 4D). Speculatively, longer RNA species could travel though the surface exposed half-channel and interconnect all components of the core complex into a highly organised assembly on this target RNA (Figure 5).

Figure 5. — When the preferred carbon source, succinate, is abundant, cellular CrcZ levels are low and Hfq and Crc occlude the *amiE* ribosome binding site by forming a higher order assembly, rendering CCR active and repressing synthesis of for example aliphatic amidase (top). We envision that translational repression may either occur on two mRNA molecules (i) or on a single mRNA molecule that comprises another Hfq binding site downstream of the first (ii). Upon depletion of succinate, CrcZ levels increase and CrcZ sequesters Hfq and Crc from *amiE*, potentially by occupying the multiple A-R-N patches on CrcZ and forming multicomponent ‘beads on a string’ (bottom). As such CCR is abolished, allowing metabolism of a secondary carbon source, for example amide conversion by aliphatic amidase.

Discussion

Many functional studies have highlighted the cooperation of global posttranscriptional regulators in controlling the fate of targeted transcripts. A major gap in our current understanding has been the lack of high-resolution structural data of these highly coordinated cellular processes. Here, we report for the first time structural information on how a partner protein can assist the role of Hfq in direct translational repression when bound to a translational initiation region (amiE_6ARN) by forming a multi-component assembly. The added value of the Crc in the complex as compared to recognition of the A-rich sequence by Hfq is the potential for highly cooperative switches for activation/repression and greater specificity.

The amiE RNA translation control site is an A-rich fragment that occupies almost entirely the distal surface of the P. aeruginosa Hfq, weaving in between basic, surface exposed islands (Figure 2). There are striking similarities to the structure of the polyA₁₈ complex with E. coli Hfq (Link et al., 2009), which greatly added to the understanding of RNA binding and chaperone mechanisms, and hinted at how the distinct polyA RNA interaction might enable Hfq-mediated regulation. The polyA₁₈/Hfq structure revealed rules for recognition of motifs of the type A-R-N, where R is purine and N is any base (Figure 2—figure supplement 1). The P. aeruginosa Hfq interaction with amiE_6ARN follows the same rules. The A-R-N repeat occurs in many RNAs, and it is a recurring motif in the nascent transcripts that associate with Hfq and Crc in Pseudomonas (Kambara et al., 2018). It has been proposed that the exposed bases (at the Entry/Exit site, corresponding to ‘N’ in the A-R-N motif) could mediate RNA to RNA interactions (Schulz et al., 2017), but we observe that the exposed bases are presented for protein recognition. The exposed bases at the ‘N’ position and RNA backbone in the Hfq/amiE_6ARN complex are available for interactions with Crc to form a cooperative assembly (Figures 3 and 4) that mediates translational repression on target mRNAs, that is carbon catabolite repression in vivo when the preferred carbon source is available. Strikingly, the A-R-N motif is not supported by the Hfq of the Gram-positive bacteria Staphylococcus aureus and Bacillus subtilis (Horstmann et al., 2012) (Figure 2—figure supplement 1B). In view of this divergence, it seems unlikely that Hfq in the Gram-positive species form assemblies that resemble the P. aeruginosa Hfq/Crc assembly.

Previous studies have shown that both Hfq and Crc are required for tight translational repression of mRNAs, which are subject to carbon catabolite repression (CCR) (Sonnleitner and Bläsi, 2014; Moreno et al., 2015). The presence of Crc did not significantly enhance the affinity of Hfq for amiE_6ARN RNA (Sonnleitner et al., 2018). However, the simultaneous interactions of Crc with both binding partners resulted in an Hfq/Crc/RNA assembly with increased stability when compared with the Hfq/RNA complex alone (Sonnleitner et al., 2018). In light of our structural studies, the enhancing effect of Crc in Hfq-mediated translational repression of target mRNAs during CCR (Sonnleitner and Bläsi, 2014; Moreno et al., 2015) can be explained by sandwiching the Hfq binding site on mRNA between both binding partners. Thus, the structural model can rationalize the observed increased stability of the Hfq/Crc/amiE_6ARN assembly when compared to the sole Hfq/amiE_6ARN complex (Sonnleitner et al., 2018). It is conceivable that full repression is only achieved when amiE_6ARN is masked entirely in the 2:4:2 complex, which is supported by our in vivo studies.

The question arises why a higher order assembly such as the 2:2:2 core is formed and not a simpler complex. The structural data indicate that the dimerization of Crc provides the key step for formation of the 2:2:2 complex, because it will pre-organise a copy of the surface that interacts with a binary Hfq/RNA complex so that a second Hfq/RNA complex can be recruited. Thus, all components seem to be necessary to form the complex so that there is no formation of lower order ‘sub assemblies’. The structural data are consistent with Crc having no capacity for RNA binding by itself (Milojevic et al., 2013). The Hfq/Crc/RNA complex may thus be assembled in a checklist-like manner through numerous small contacting surfaces and when the RNA target is presented by Hfq in a specific, well-defined configuration. In this way, the components interact mutually through chelate cooperative effects. It seem likely that the 2:2:2 core is formed first followed by recruitment of the other Crc components.

We envisage that the 2:2:2 core and higher order assemblies might interact with other mRNAs. The higher order assembly could capture two of such mRNA substrates as shown in Figure 5i, but chelate effects might instead induce formation of the complex on a single mRNA target. In that scenario, a portion of the mRNA could thread through the central basic half channel (Figure 4D) as depicted in Figure 5ii and potentially recruit the second Hfq hexamer. In this context it is worth noting that several mRNAs, including amiE, which are directly repressed by Hfq/Crc comprise another putative Hfq binding motif downstream of the start codon (Sonnleitner et al., 2018). Therefore, we are currently exploring the possibility whether this second Hfq binding motif contributes to Hfq/Crc assembly on longer mRNA substrates and whether it likewise may confer specificity for these Hfq/Crc repressed substrates.

Under conditions of catabolite repression regulation, pull-down assays showed that Hfq and Crc form a co-complex in the presence of the 426nt long CrcZ RNA (Moreno et al., 2015; Sonnleitner et al., 2018). In the presence of less preferred carbon sources, the expression levels of CrcZ RNA increase (Sonnleitner et al., 2009) and CrcZ functions as an antagonist in Hfq/Crc mediated translational repression of catabolic genes. The CrcZ RNA has multiple A-R-N triplets that could be sites for Hfq/Crc interaction (Sonnleitner and Bläsi, 2014) and sequester multiple Hfq/Crc proteins (Figure 5). Thus, under conditions where CCR is relieved, CrcZ RNA would serve as a sponge for Hfq/Crc to prevent repression of genes encoding proteins required for the utilization of less preferred carbon sources (Figure 5). How the CrcZ RNA is displaced from Hfq/Crc remains unknown. However, the assemblies are likely to be dynamic and the displacement process might resemble that proposed for the step-wise exchange of sRNAs on Hfq (Fender et al., 2010).

Recent findings show that the regulatory spectrum of Hfq and Crc is much broader than initially expected. Hfq was found to bind more than 600 nascent transcripts co-transcriptionally often in concert with Crc (Kambara et al., 2018). These findings indicate that Hfq and Crc together regulate gene expression post-transcriptionally beyond just catabolite repression. Understanding how gene expression is regulated post-transcriptionally in pathogens such as P. aeruginosa may provide potential targets for novel drug design. Hfq and Crc are involved in key metabolic and virulence processes in Pseudomonas species (O'Toole et al., 2000; Sonnleitner et al., 2003; Sonnleitner et al., 2006; Linares et al., 2010; Huang et al., 2012; Zhang et al., 2012; Zhang et al., 2013; Sonnleitner and Bläsi, 2014; Pusic et al., 2016). Disrupting the interface of the core assembly of the Hfq/Crc complex might be one strategy to counter, among other, metabolic regulation and consequently its downstream processes that impact on virulence during infection. A recent study showed how overproduction of the aliphatic amidase AmiE strongly reduced biofilm formation and almost fully attenuated virulence in, amongst others, a mouse model of acute lung infection (Clamens et al., 2017). Novel drugs that specifically counteract Hfq/Crc/amiE assembly formation and prevent repression of AmiE production could induce the phenotype described by Clamens et al. (2017). The high resolution structures presented here provide a starting point for novel strategies to interfere with carbon regulation in a pathogenic bacterium for therapeutic intervention of threatening infections.

Materials and methods

Key resources table.

Reagent type (species) or resource	Designation	Source or reference	Identifiers	Additional information
Gene (Pseudomonas aeruginosa)	crc	NA	Pseudomonas genome database: PA5332
Gene (P. aeruginosa)	amiE	NA	Pseudomonas genome database: PA3366
Strain, strain background (P. aeruginosa)	PAO1	PMID: 111024
Strain, strain background (P. aeruginosa)	PAO1Δcrc	PMID: 20080802
Strain, strain background (Escherichia coli)	XL1-Blue	Stratagene
Strain, strain background (E. coli)	BL21(DE3)	Novagen
Genetic reagent (P. aeruginosa)	amiE_6ARN RNA	Microsynth, PMID: 29244160	sequence: 5'-AAAAAUAACAACAAGAAG-3'
Antibody	anti-Crc (rabbit polyclonal)	Pineda	(1:5000)
Antibody	anti-Hfq (rabbit polyclonal)	Pineda	(1:10000)
Antibody	anti-S1 (rabbit polyclonal)	Pineda	(1:10000)
Antibody	anti-rabbit IgG (goat polyclonal)	Sigma	Sigma: A3687	conjugated with alkaline phosphatase, (1:10000)
Recombinant DNA reagent	pME9655 (plasmid)	PMID: 20080802
Recombinant DNA reagent	pETM14lic-His₆Crc (plasmid)	PMID: 23717639
Recombinant DNA reagent	pME4510crc_Flag (plasmid)	PMID: 29244160
Peptide, recombinant protein	Hfq	PMID: 21330354
Peptide, recombinant protein	Crc	PMID: 23717639
Software, algorithm	Relion3	DOI: 10.7554/eLife.42166
Software, algorithm	LocScale	DOI: 10.7554/eLife.27131
Software, algorithm	LocalDeblur	DOI: 10.1101/433284

Open in a new tab

Protein synthesis, purification and complex formation

P. aeruginosa Hfq and Crc were produced in E. coli and purified as described by Sonnleitner et al. (2018). The synthetic 18-mer amiE_6ARN RNA (5´-AAAAAUAACAACAAGAGG-3´) used in these studies consists of six tripartite binding motifs (Sonnleitner and Bläsi, 2014). The Hfq/Crc/RNA complex was prepared by first heating the amiE_6ARN RNA at 95°C for 5 min followed by 50°C for 10 min and 37°C for 10 min. The RNA was then incubated with the Hfq hexamer at a 1:1 molar ratio on ice for 20 min to form a binary complex, then an equal molar ratio of Crc was added as recently described (Sonnleitner et al., 2018). The mixture was incubated on ice for 30 min prior to fractionation by size exclusion chromatography using a Superdex 200 column equilibrated in running buffer composed of 20 mM HEPES, pH 7.9, 10 mM KCl, 40 mM NaCl, 1 mM MgCl₂, and 2 mM TCEP (tris(2-carboxyethyl) phosphine). The peak fractions were buffer exchanged into 20 mM HEPES, pH 7.9, 10 mM KCl, 40 mM NaCl, 5 mM MgCl₂. Samples used for cross-linking were incubated with bis(sulfosuccinimidyl)suberate (BS³) at 150 µM for 30 min on ice, followed by quenching at 37.5 mM Tris-HCl pH 8.0.

CryoEM specimen preparation and data acquisition

Graphene oxide grids are prepared as described by Pantelic et al. (2010). Briefly, 2 mg/ml of graphene oxide solution in water (Aldrich) was diluted ten times in water. After removing aggregation by spinning for 30 s at 300 rcf, 2 µl of graphene oxide solution was loaded on freshly glow discharged quantifoil Au-grids (R1.2/1.3, 300 mesh). Glow discharge was performed prior to graphene oxide coating at 45 mA for 60 s with an Edward Sputter Coater S150B at 0.2 m Bar at 0.75 KV. After the graphene oxide had been adsorbed for 1 min, the grids were washed 3 times with 20 μl water, then air-dried for 1 hr at room temperature prior to sample application. Specimens for cryoEM analysis were prepared by applying 2 μl of a 0.65 µM solution of the Hfq/Crc/RNA complex to the Quantifoil Au grids freshly coated with graphene oxide. After an adsorption time of 60 s, the grids were blotted for 10 s at a blot force of 5, then plunge frozen into liquid ethane using a Vitrobot (FEI). Images were recorded on a Krios G2, Falcon III direct electron detector at 300 kV operating in counting mode (Supplementary file 1C).

Movie processing, single particle analysis, 3D reconstruction and refinement

Whole frame motion correction was performed on movies with motioncorr2 with dose weighting followed by CTF estimation using gctf (Zhang, 2016; Zheng et al., 2017). RELION-3.0 was used for data processing (Scheres, 2012). Final resolution estimates were calculated after the application of a soft binary mask and phase randomisation and determined based on the gold standard FSC = 0.143 criterion (Scheres and Chen, 2012; Chen et al., 2013).

For the BS³ treated complex, after manually picking 3159 particles and using suitable references for autopicking, 482426 particles were used for early classifications. After three rounds of rejecting particles by 2D classification, 215774 particles were used for initial model generation and 3D classification. An initial model was generated using an SGD algorithm based on a small subset of particles with diverse orientations (Punjani et al., 2017). During 3D classification, three different complexes were resolved after 25 iterations with an angular sampling of 7.5°: 2Hfq:2Crc:2amiE_6ARN (2:2:2), 2Hfq:3Crc:2amiE_6ARN (2:3:2) and 2Hfq:4Crc:2amiE_6ARN (2:4:2). To properly separate, validate and refine the three classes, the same 3D classification was rerun with the new 2:3:2 model as reference model, lowpass filtered to 20 Å resolution. C2 symmetry was observed and imposed for the 2:2:2 and 2:4:2 complexes. Each of the classes was then refined to sub-3.5 Å resolution, followed by per-particle frame alignment for movement correction and per-frame damage weighting. The resulting ‘polished’ particles were subjected to a final refinement round with solvent flattening. All reference models were lowpass filtered to 60 Å prior to refinement. The dominant class (2:2:2) had a resolution of 3.13 Å. Local resolution calculations were done with the RELION local resolution estimator, Supplementary file 1A (Figure 1—figure supplement 1). The 2:4:2 and 2:3:2 maps were sharpened locally with LocScale and LocalDeblur to better resolve the additional Crc components (Jakobi et al., 2017; Ramírez-aportela et al., 2018).

Crystal structures for P. aeruginosa Crc (PDB code 1U1S) and Hfq (PDB code 4JG3) were manually docked into the EM density map as rigid bodies in Chimera (Pettersen et al., 2004). The RNA 18-mers were manually built into the density using Coot (Emsley et al., 2010). Refmac5 and Phenix real-space refinement with global energy minimization, NCS-restraints, group B-factor and geometry restraints were used to iteratively refine the multi-subunit complexes at high resolution, followed by manual corrections for Ramachandran and geometric outliers in Coot and ISOLDE (Supplementary file 1A) (Emsley et al., 2010; Murshudov et al., 2011; Afonine et al., 2012; Croll, 2018). Model quality was evaluated with Procheck in CCP4 and MolProbity (Williams et al., 2018). In silico 2 Å maps were generated from the atomic models and FSC validation against the experimental maps was performed with the EMDB Fourier shell correlation server (EMBL-EBI) (Figure 1—figure supplement 2B).

Bacterial strains and plasmids

The strains, plasmids and oligonucleotides used in this study are listed in Supplementary file 1B and 1C.

Construction of plasmids encoding Crc variant proteins for in vivo translational repression assay

To test the proficiency of Crc mutant proteins to co-repress translation of a translational amiE:lacZ reporter gene, derivatives of plasmid pME4510crc_Flag (Supplementary file 1B) were constructed by means of Quick change site directed mutagenesis (Agilent Technologies). Plasmid pME4510crc_Flag was used together with the corresponding mutagenic oligonucleotide pairs (Supplementary file 1C). The parental plasmid templates were digested with DpnI and the mutated nicked circular strands were transformed into E. coli XL1-Blue, generating plasmids pME4510crc_(R140E)Flag, pME4510crc_(E142R)Flag, pME4510crc_(R229E)Flag, pME4510crc_(E193R)Flag, pME4510crc_(R230E)Flag, pME4510crc_{(E142R,R229E)Flag}, pME4510crc_{(E193R,R230E)Flag}, pME4510crc_{(E193A,R230E)Flag}, pME4510crc_{(E142R,R230E)Flag} and pME4510crc_{(E142R, R229E, R230E)Flag} and pME4510crc_{(E142A, R229E, R230E)Flag}.

In vivo translational repression of an amiE::lacZ reporter gene in the presence of Crc variants

The ability of the Crc mutant proteins to repress translation of an amiE::lacZ reporter gene was tested in a PAO1 crc deletion strain bearing plasmids encoding the wt protein or the respective Crc variants (Supplementary file 1B) as described by Sonnleitner et al. (2018). The β-galactosidase activities were determined as described (Miller, 1972). The β-galactosidase units in the different experiments were derived from two independent experiments.

Construction of plasmids employed for the production of selected Crc mutant proteins

The R140E, E142R, R230E single aa exchanges in Crc were obtained by using the QuickChange site-directed mutagenesis protocol (Agilent Technologies). The plasmid pETM14lic-His₆Crc (Supplementary file 1B) was used together with the corresponding mutagenic oligonucleotide pairs (Supplementary file 1C). The entire plasmids were amplified with Pfu DNA polymerase (Thermo Scientific). The parental plasmid templates were digested with DpnI and the mutated nicked circular strands were transformed into E. coli XL1-Blue, generating plasmids pETM14lic-His6Crc_R140E, pETM14lic-His6Crc_E142R and pETM14lic-His6Crc_R230E.

Purification of Crc and Crc variants

The Crc protein and the Crc variants Crc_R140E, Crc_E142R and Crc_R230E were purified from E. coli strain BL21(DE3) harboring either plasmid pETM14lic-His Crc or the respective derivatives using Ni-affinity chromatography, followed by removal of the His₆-tag with GST-HRV14-3C ‘‘PreScission’’ protease as described by Milojevic et al., 2013.

In vitro co-IP studies

The co-IP studies in the presence of 40 pmol of Hfq-hexamer, 120 pmol of Crc protein or of the respective Crc mutant proteins and 40 pmol amiE_6ARN RNA were performed as described (Sonnleitner et al., 2018).

Western blot analyses

Equal amounts of proteins were separated on 12% SDS-polyacrylamide gels, and then electro-blotted onto a nitrocellulose membrane. The blots were blocked with 5% dry milk in TBS buffer, and probed with rabbit anti-Hfq (Pineda) and rabbit anti-Crc (Pineda) antibodies, respectively. Immuno-detection of ribosomal protein S1 served as a loading control. The antibody-antigen complexes were visualized with alkaline-phosphatase conjugated secondary antibodies (Sigma) using the chromogenic substrates nitro blue tetrazolium chloride (NBT) and 5-Bromo-4-chloro-3-indolyl phosphate (BCIP).

Acknowledgements

The coordinates and cryoEM maps have been deposited in the PDB and the EMBD. BFL, XYP and TD are supported by the Welcome Trust (200873/Z/16/Z). TD is also supported by an AstraZeneca Studentship. UB and ES are supported by the Austrian Science Fund (FWF) (www.fwf.ac.at/en) [P28711-B22]. We thank our colleagues Jamie Blaza, Dima Chirgadze, Jiri Sponer, Miroslav Kreply, Kasia Bandyra, Steven Hardwick, Sjors Scheres, Joerg Vogel, Armin Resch and Nguyen Thi Bach Hue for advice, helpful discussions and support. For access and help at facilities, we thank Giuseppe Cannon and staff at the MRC-LMB EM Facility and Kasim Sader at Thermo Fisher Scientific Pharma CryoEM Facility, Nanoscience Centre of University of Cambridge.

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

Udo Bläsi, Email: udo.blaesi@univie.ac.at.

Ben F Luisi, Email: bfl20@cam.ac.uk.

Anna Marie Pyle, Yale University, United States.

John Kuriyan, University of California, Berkeley, United States.

Funding Information

This paper was supported by the following grants:

Wellcome Trust 200873/Z/16/Z to Ben F Luisi, Xue Yuan Pei, Tom Dendooven.
AstraZeneca Studentship to Tom Dendooven.
Austrian Science Fund P28711-B22 to Udo Bläsi, Elisabeth Sonnleitner.

Additional information

Competing interests

No competing interests declared.

Author contributions

Data curation, Formal analysis, Investigation.

Conceptualization, Data curation, Formal analysis, Investigation, Methodology.

Data curation, Formal analysis.

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

Conceptualization, Funding acquisition, Investigation, Project administration.

Additional files

Supplementary file 1. Resolution of maps and angular distributions of 2D images.

(A) Summary of the data collection parameters and the reconstruction and refinement statistics for the Hfq:Crc:RNA complexes. (B) List of the plasmids and strains used in this study. (C) Oligonucleotides used in this study.

elife-43158-supp1.docx^{(39.6KB, docx)}

DOI: 10.7554/eLife.43158.013

Transparent reporting form

elife-43158-transrepform.docx^{(246.6KB, docx)}

DOI: 10.7554/eLife.43158.014

Data availability

CryoEM data have been deposited in the Electron Microscopy Data Bank.

The following datasets were generated:

Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Blasi, Ben F Luisi. 2019. Hfq/Crc/RNA cpmplex 2:4:2 ratio. Protein Data Bank in Europe. 6o1m

Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Blasi, Ben F Luisi. 2019. Hfq/Crc/RNA complex 2:2:2 ratio. Electron Microscopy Data Bank. EMD-0590

Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Blasi, Ben F Luisi. 2019. Hfq/Crc/RNA complex 2:3:2 ratio. Electron Microscopy Data Bank. EMD-0591

Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Blasi, Ben F Luisi. 2019. Hfq/Crc/RNA complex 2:4:2 ratio. Electron Microscopy Data Bank. EMD-0592

Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Bläsi, Ben F Luisi. 2019. Hfq/Crc/RNA complex 2:2:2 ratio. Protein Data Bank in Europe. 6o1k

Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Bläsi. 2019. Hfq/Crc/RNA complex 2:3:2 ratio. Protein Data Bank in Europe. 6o1l

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]
Bandyra KJ, Sinha D, Syrjanen J, Luisi BF, De Lay NR. The ribonuclease polynucleotide phosphorylase can interact with small regulatory RNAs in both protective and degradative modes. RNA. 2016;22:360–372. doi: 10.1261/rna.052886.115. [DOI] [PMC free article] [PubMed] [Google Scholar]
Chen S, McMullan G, Faruqi AR, Murshudov GN, Short JM, Scheres SH, Henderson R. High-resolution noise substitution to measure overfitting and validate resolution in 3D structure determination by single particle electron cryomicroscopy. Ultramicroscopy. 2013;135:24–35. doi: 10.1016/j.ultramic.2013.06.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
Clamens T, Rosay T, Crépin A, Grandjean T, Kentache T, Hardouin J, Bortolotti P, Neidig A, Mooij M, Hillion M, Vieillard J, Cosette P, Overhage J, O'Gara F, Bouffartigues E, Dufour A, Chevalier S, Guery B, Cornelis P, Feuilloley MG, Lesouhaitier O. The aliphatic amidase AmiE is involved in regulation of pseudomonas aeruginosa virulence. Scientific Reports. 2017;7:41178. doi: 10.1038/srep41178. [DOI] [PMC free article] [PubMed] [Google Scholar]
Croll TI. ISOLDE : a physically realistic environment for model building into low-resolution electron-density maps. Acta Crystallographica Section D Structural Biology. 2018;74:519–530. doi: 10.1107/S2059798318002425. [DOI] [PMC free article] [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]
Fender A, Elf J, Hampel K, Zimmermann B, Wagner EGH. RNAs actively cycle on the Sm-like protein hfq. Genes & Development. 2010;24:2621–2626. doi: 10.1101/gad.591310. [DOI] [PMC free article] [PubMed] [Google Scholar]
Fernández L, Breidenstein EB, Taylor PK, Bains M, de la Fuente-Núñez C, Fang Y, Foster LJ, Hancock RE. Interconnection of post-transcriptional regulation: the RNA-binding protein hfq is a novel target of the lon protease in Pseudomonas aeruginosa. Scientific Reports. 2016;6:26811. doi: 10.1038/srep26811. [DOI] [PMC free article] [PubMed] [Google Scholar]
Heitzinger DA. University of Vienna; 2016. Impact of Hfq on Antibiotic Susceptibility of Pseudomonas Aeruginosa. Master Thesis. [Google Scholar]
Holloway BW, Krishnapillai V, Morgan AF. Chromosomal genetics of pseudomonas. Microbiological Reviews. 1979;43:73–102. doi: 10.1128/mr.43.1.73-102.1979. [DOI] [PMC free article] [PubMed] [Google Scholar]
Horstmann N, Orans J, Valentin-Hansen P, Shelburne SA, Brennan RG. Structural mechanism of staphylococcus aureus Hfq binding to an RNA A-tract. Nucleic Acids Research. 2012;40:11023–11035. doi: 10.1093/nar/gks809. [DOI] [PMC free article] [PubMed] [Google Scholar]
Huang J, Sonnleitner E, Ren B, Xu Y, Haas D. Catabolite repression control of pyocyanin biosynthesis at an intersection of primary and secondary metabolism in pseudomonas aeruginosa. Applied and Environmental Microbiology. 2012;78:5016–5020. doi: 10.1128/AEM.00026-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ikeda Y, Yagi M, Morita T, Aiba H. Hfq binding at RhlB-recognition region of RNase E is crucial for the rapid degradation of target mRNAs mediated by sRNAs in Escherichia coli. Molecular Microbiology. 2011;79:419–432. doi: 10.1111/j.1365-2958.2010.07454.x. [DOI] [PubMed] [Google Scholar]
Jakobi AJ, Wilmanns M, Sachse C. Model-based local density sharpening of cryo-EM maps. eLife. 2017;6:e27131. doi: 10.7554/eLife.27131. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kambara TK, Ramsey KM, Dove SL. Pervasive targeting of nascent transcripts by hfq. Cell Reports. 2018;23:1543–1552. doi: 10.1016/j.celrep.2018.03.134. [DOI] [PMC free article] [PubMed] [Google Scholar]
Linares JF, Moreno R, Fajardo A, Martínez-Solano L, Escalante R, Rojo F, Martínez JL. The global regulator crc modulates metabolism, susceptibility to antibiotics and virulence in Pseudomonas aeruginosa. Environmental Microbiology. 2010;12:3196–3212. doi: 10.1111/j.1462-2920.2010.02292.x. [DOI] [PubMed] [Google Scholar]
Link TM, Valentin-Hansen P, Brennan RG. Structure of Escherichia coli hfq bound to polyriboadenylate RNA. PNAS. 2009;106:19292–19297. doi: 10.1073/pnas.0908744106. [DOI] [PMC free article] [PubMed] [Google Scholar]
Lu P, Wang Y, Zhang Y, Hu Y, Thompson KM, Chen S. RpoS-dependent sRNA RgsA regulates fis and AcpP in Pseudomonas aeruginosa. Molecular Microbiology. 2016;102:244–259. doi: 10.1111/mmi.13458. [DOI] [PubMed] [Google Scholar]
Miller JH. Experiments in Molecular Genetics. New York: Cold Spring Harbor Press; 1972. [Google Scholar]
Milojevic T, Grishkovskaya I, Sonnleitner E, Djinovic-Carugo K, Bläsi U. The Pseudomonas aeruginosa catabolite repression control protein Crc is devoid of RNA binding activity. PLOS ONE. 2013;8:e64609. doi: 10.1371/journal.pone.0064609. [DOI] [PMC free article] [PubMed] [Google Scholar]
Mohanty BK, Maples VF, Kushner SR. The Sm-like protein hfq regulates polyadenylation dependent mRNA decay in Escherichia coli. Molecular Microbiology. 2004;54:905–920. doi: 10.1111/j.1365-2958.2004.04337.x. [DOI] [PubMed] [Google Scholar]
Moreno R, Hernández-Arranz S, La Rosa R, Yuste L, Madhushani A, Shingler V, Rojo F. The crc and hfq proteins of Pseudomonas putida cooperate in catabolite repression and formation of ribonucleic acid complexes with specific target motifs. Environmental Microbiology. 2015;17:105–118. doi: 10.1111/1462-2920.12499. [DOI] [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]
O'Toole GA, Gibbs KA, Hager PW, Phibbs PV, Kolter R. The global carbon metabolism regulator crc is a component of a signal transduction pathway required for biofilm development by pseudomonas aeruginosa. Journal of Bacteriology. 2000;182:425–431. doi: 10.1128/JB.182.2.425-431.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
Panja S, Schu DJ, Woodson SA. Conserved arginines on the rim of hfq catalyze base pair formation and exchange. Nucleic Acids Research. 2013;41:7536–7546. doi: 10.1093/nar/gkt521. [DOI] [PMC free article] [PubMed] [Google Scholar]
Pantelic RS, Meyer JC, Kaiser U, Baumeister W, Plitzko JM. Graphene oxide: a substrate for optimizing preparations of frozen-hydrated samples. Journal of Structural Biology. 2010;170:152–156. doi: 10.1016/j.jsb.2009.12.020. [DOI] [PubMed] [Google Scholar]
Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, Ferrin TE. UCSF chimera--a visualization system for exploratory research and analysis. Journal of Computational Chemistry. 2004;25:1605–1612. doi: 10.1002/jcc.20084. [DOI] [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]
Pusic P, Tata M, Wolfinger MT, Sonnleitner E, Häussler S, Bläsi U. Cross-regulation by CrcZ RNA controls anoxic biofilm formation in pseudomonas aeruginosa. Scientific Reports. 2016;6:39621. doi: 10.1038/srep39621. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ramírez-aportela E, Vilas JL, Melero R, Conesa P. Automatic local resolution-based sharpening of cryo-EM maps. bioRxiv. 2018 doi: 10.1101/433284. [DOI] [PMC free article] [PubMed]
Resch A, Većerek B, Palavra K, Bläsi U. Requirement of the CsdA DEAD-box helicase for low temperature riboregulation of rpoS mRNA. RNA Biology. 2010;7:796–802. doi: 10.4161/rna.7.6.13768. [DOI] [PMC free article] [PubMed] [Google Scholar]
Rojo F. Carbon catabolite repression in Pseudomonas : optimizing metabolic versatility and interactions with the environment. FEMS Microbiology Reviews. 2010;34:658–684. doi: 10.1111/j.1574-6976.2010.00218.x. [DOI] [PubMed] [Google Scholar]
Salvail H, Caron MP, Bélanger J, Massé E. Antagonistic functions between the RNA chaperone hfq and an sRNA regulate sensitivity to the antibiotic colicin. The EMBO Journal. 2013;32:2764–2778. doi: 10.1038/emboj.2013.205. [DOI] [PMC free article] [PubMed] [Google Scholar]
Santiago-Frangos A, Kavita K, Schu DJ, Gottesman S, Woodson SA. C-terminal domain of the RNA chaperone hfq drives sRNA competition and release of target RNA. PNAS. 2016;113:E6089–E6096. doi: 10.1073/pnas.1613053113. [DOI] [PMC free article] [PubMed] [Google Scholar]
Santiago-Frangos A, Jeliazkov JR, Gray JJ, Woodson SA. Acidic C-terminal domains autoregulate the RNA chaperone hfq. eLife. 2017;6:e27049. doi: 10.7554/eLife.27049. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sauer E, Schmidt S, Weichenrieder O. Small RNA binding to the lateral surface of hfq hexamers and structural rearrangements upon mRNA target recognition. PNAS. 2012;109:9396–9401. doi: 10.1073/pnas.1202521109. [DOI] [PMC free article] [PubMed] [Google Scholar]
Scheres SH. A bayesian view on cryo-EM structure determination. Journal of Molecular Biology. 2012;415:406–418. doi: 10.1016/j.jmb.2011.11.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
Scheres SH, Chen S. Prevention of overfitting in cryo-EM structure determination. Nature Methods. 2012;9:853–854. doi: 10.1038/nmeth.2115. [DOI] [PMC free article] [PubMed] [Google Scholar]
Schulz EC, Seiler M, Zuliani C, Voigt F, Rybin V, Pogenberg V, Mücke N, Wilmanns M, Gibson TJ, Barabas O. Intermolecular base stacking mediates RNA-RNA interaction in a crystal structure of the RNA chaperone hfq. Scientific Reports. 2017;7:9903. doi: 10.1038/s41598-017-10085-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Schumacher MA, Pearson RF, Møller T, Valentin-Hansen P, Brennan RG. Structures of the pleiotropic translational regulator hfq and an Hfq-RNA complex: a bacterial Sm-like protein. The EMBO Journal. 2002;21:3546–3556. doi: 10.1093/emboj/cdf322. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sonnleitner E, Hagens S, Rosenau F, Wilhelm S, Habel A, Jäger KE, Bläsi U. Reduced virulence of a hfq mutant of pseudomonas aeruginosa O1. Microbial Pathogenesis. 2003;35:217–228. doi: 10.1016/S0882-4010(03)00149-9. [DOI] [PubMed] [Google Scholar]
Sonnleitner E, Schuster M, Sorger-Domenigg T, Greenberg EP, Bläsi U. Hfq-dependent alterations of the transcriptome profile and effects on quorum sensing in pseudomonas aeruginosa. Molecular Microbiology. 2006;59:1542–1558. doi: 10.1111/j.1365-2958.2006.05032.x. [DOI] [PubMed] [Google Scholar]
Sonnleitner E, Abdou L, Haas D. Small RNA as global regulator of carbon catabolite repression in pseudomonas aeruginosa. PNAS. 2009;106:21866–21871. doi: 10.1073/pnas.0910308106. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sonnleitner E, Wulf A, Campagne S, Pei XY, Wolfinger MT, Forlani G, Prindl K, Abdou L, Resch A, Allain FH, Luisi BF, Urlaub H, Bläsi U. Interplay between the catabolite repression control protein crc, hfq and RNA in Hfq-dependent translational regulation in pseudomonas aeruginosa. Nucleic Acids Research. 2018;46:1470–1485. doi: 10.1093/nar/gkx1245. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sonnleitner E, Bläsi U. Regulation of hfq by the RNA CrcZ in Pseudomonas aeruginosa carbon catabolite repression. PLOS Genetics. 2014;10:e1004440. doi: 10.1371/journal.pgen.1004440. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sukhodolets MV, Garges S. Interaction of Escherichia coli RNA polymerase with the ribosomal protein S1 and the Sm-like ATPase Hfq. Biochemistry. 2003;42:8022–8034. doi: 10.1021/bi020638i. [DOI] [PubMed] [Google Scholar]
Valentini M, García-Mauriño SM, Pérez-Martínez I, Santero E, Canosa I, Lapouge K. Hierarchical management of carbon sources is regulated similarly by the CbrA/B systems in pseudomonas aeruginosa and Pseudomonas putida. Microbiology. 2014;160:2243–2252. doi: 10.1099/mic.0.078873-0. [DOI] [PubMed] [Google Scholar]
Van den Bossche A, Ceyssens PJ, De Smet J, Hendrix H, Bellon H, Leimer N, Wagemans J, Delattre AS, Cenens W, Aertsen A, Landuyt B, Minakhin L, Severinov K, Noben JP, Lavigne R. Systematic identification of hypothetical bacteriophage proteins targeting key protein complexes of pseudomonas aeruginosa. Journal of Proteome Research. 2014;13:4446–4456. doi: 10.1021/pr500796n. [DOI] [PubMed] [Google Scholar]
Vecerek B, Moll I, Bläsi U. Translational autocontrol of the Escherichia coli hfq RNA chaperone gene. RNA. 2005;11:976–984. doi: 10.1261/rna.2360205. [DOI] [PMC free article] [PubMed] [Google Scholar]
Vecerek B, Beich-Frandsen M, Resch A, Bläsi U. Translational activation of rpoS mRNA by the non-coding RNA DsrA and hfq does not require ribosome binding. Nucleic Acids Research. 2010;38:1284–1293. doi: 10.1093/nar/gkp1125. [DOI] [PMC free article] [PubMed] [Google Scholar]
Vogel J, Luisi BF. Hfq and its constellation of RNA. Nature Reviews Microbiology. 2011;9:578–589. doi: 10.1038/nrmicro2615. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wagner EGH, Romby P. Small RNAs in bacteria and archaea: who they are, what they do, and how they do it. Advances in Genetics. 2015;90:133–208. doi: 10.1016/bs.adgen.2015.05.001. [DOI] [PubMed] [Google Scholar]
Wei Y, Zhang H, Gao ZQ, Xu JH, Liu QS, Dong YH. Structure analysis of the global metabolic regulator crc from Pseudomonas aeruginosa. IUBMB Life. 2013;65:50–57. doi: 10.1002/iub.1103. [DOI] [PubMed] [Google Scholar]
Williams CJ, Headd JJ, Moriarty NW, Prisant MG, Videau LL, Deis LN, Verma V, Keedy DA, Hintze BJ, Chen VB, Jain S, Lewis SM, Arendall WB, Snoeyink J, Adams PD, Lovell SC, Richardson JS, Richardson DC. MolProbity: more and better reference data for improved all-atom structure validation. Protein Science. 2018;27:293–315. doi: 10.1002/pro.3330. [DOI] [PMC free article] [PubMed] [Google Scholar]
Worrall JA, Górna M, Crump NT, Phillips LG, Tuck AC, Price AJ, Bavro VN, Luisi BF. Reconstitution and analysis of the multienzyme Escherichia coli RNA degradosome. Journal of Molecular Biology. 2008;382:870–883. doi: 10.1016/j.jmb.2008.07.059. [DOI] [PMC free article] [PubMed] [Google Scholar]
Yang N, Ding S, Chen F, Zhang X, Xia Y, Di H, Cao Q, Deng X, Wu M, Wong CC, Tian XX, Yang CG, Zhao J, Lan L. The crc protein participates in down-regulation of the lon gene to promote rhamnolipid production and rhl quorum sensing in pseudomonas aeruginosa. Molecular Microbiology. 2015;96:526–547. doi: 10.1111/mmi.12954. [DOI] [PubMed] [Google Scholar]
Zhang L, Chiang WC, Gao Q, Givskov M, Tolker-Nielsen T, Yang L, Zhang G. The catabolite repression control protein Crc plays a role in the development of antimicrobial-tolerant subpopulations in Pseudomonas aeruginosa biofilms. Microbiology. 2012;158:3014–3019. doi: 10.1099/mic.0.061192-0. [DOI] [PubMed] [Google Scholar]
Zhang L, Gao Q, Chen W, Qin H, Hengzhuang W, Chen Y, Yang L, Zhang G. Regulation of pqs quorum sensing via catabolite repression control in pseudomonas aeruginosa. Microbiology. 2013;159:1931–1936. doi: 10.1099/mic.0.066266-0. [DOI] [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 JP, 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]
Zivanov J, Nakane T, Forsberg BO, Kimanius D, Hagen WJ, Lindahl E, Scheres SH. New tools for automated high-resolution cryo-EM structure determination in RELION-3. eLife. 2018;7:e42166. doi: 10.7554/eLife.42166. [DOI] [PMC free article] [PubMed] [Google Scholar]

eLife. doi: 10.7554/eLife.43158.028

Decision letter

Editor: Anna Marie Pyle¹

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

Thank you for submitting your article "Architectural principles for Hfq/Crc-mediated regulation of gene expression" for consideration by eLife. Your article has been reviewed by three peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by John Kuriyan as the Senior Editor. The reviewers have opted to remain anonymous.

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:

The authors present an exceptionally significant set of structures that reveal how Hfq and associated cofactor proteins interact with mRNA in order to facilitate translation and coordinate interactions with sRNA molecules. This is really the first glimpse of these important complexes and the molecular recognition and assembly pathways that are in play. It is truly exceptional work and the reviewers were very enthusiastic. That said, there is a need to restructure the paper so that it is focused on the most significant issues, and to provide stronger functional data to support the structural findings, as described below and in the reviews.

Essential revisions:

1) In the reviews and in subsequent conversation among the reviewers, there was unanimity on the need to restructure the paper so that it is focused on the main complexes and their significance to biology. The paper should be organized in a more effective way (please refer to the comments) and that major findings should be presented and discussed in light of their importance and relevance to the field at large. The text needs to be rewritten for clarity and context.

2) There was also general agreement that the supporting biochemical data should be strengthened, and the experiments suggested in the major comments by the referees should be conducted in some form in order to support the work. While the reviewers do not want to place undue burden on the investigators, there are places in the manuscript where critical supporting functional experiments still need to be included, as indicated. In addition, more attention must be devoted to a description of the methodological approaches, as these impact the findings directly in this case.

Reviewer #1:

This is a fascinating manuscript that not only provides new insights into the molecular mechanism of Hfq function and its interplay with cofactors, but also contains lessons for understanding regulatory RNA-protein complexes in general. The structures themselves are intriguing and they provide a trove of information on molecular recognition strategies within RNA-protein complexes. For example, Figure 2 is particularly interesting as no single type of interaction is completely new, but the way they are used on concert is extremely informative. The one weakness of the paper (in addition to a relatively poor presentation of the different types of complexes) is the weak supporting data on the protein-protein interactions. Attention to the points below would strengthen the paper and enhance its impact.

1) Subsection “Protein synthesis, purification and complex formation”: More experimental details about complex assembly are useful but also important for mechanistic understanding. Was it necessary to incubate the RNA and the Hfq hexamer first before adding Crc? What order of addition experiments were done to arrive at this procedure? How did the authors determine the stoichiometry for Crc? Combining the RNA/Hfq/Crc in a 1:1:1 ratio is hard to understand given the ultimate stoichiometries observed in the particles – so why didn't the authors go back and try again? This is important, as it is germane to the biological relevance described later in the paper. Given the distribution of particle stoichiometries (Figure 1), why were other assembly conditions not tried and if they were, why were they not used? Some of this is discussed in Sonnleitner, and specific results from that paper could be cited more clearly in support of the present manuscript.

2) Subsection “An ensemble of Hfq/Crc/amiE_6ARNRNA assemblies”, first paragraph: Mild crosslinking conditions were purportedly used. Were others tried? The end of the paragraph says crosslinking does not affect structure. So why was it used? What did it help?

3) Subsection “An ensemble of Hfq/Crc/amiE_6ARNRNA assemblies”, last paragraph: In the first part of the paper, there is too little focus on the overall organization and features of the 2:2:2 complex and far too much focus on the organization of complexes in the different stoichiometries – it distracts from the basic elements that are common to all three and which are likely to be of central importance. For example, the discussion of symmetry is interesting, but not central and will not help with general understanding of function by most readers. Instead, in this part of the paper, it would be good to focus on overall organization of the 2:2:2 given that it is later suggested to be the most relevant one. In the opinion of this reviewer, the first part of the paper should not have figures and text with such an emphasis on alternative stoichiometries.

4) The pull-down experimental data shown in Figure 2—figure supplement 1 are qualitative and not well explained. It's not clear to the reader how the raw data demonstrate complex formation or specific interfaces unambiguously, at least as shown. Given the affinities known for these proteins, using EMSA or size/exclusion+MALS as experimental techniques on mutants would significantly strengthen this paper. That said, a clear and compelling explanation of why the existing data are sufficient might obviate the need for these experiments.

5) Subsection “Function, origins and validation of subunit cooperativity in the 2:2:2 complex”, third paragraph. The authors state that mutant Crc_R140E did not co-IP with Hfq – but the data in the Figure 2—figure supplement 1B look just like WT Crc, so this seems to undermine the authors' point, unless there is some type of misunderstanding. Another issue: it appears from the structure that R140 also participates in the Crc1-Crc2 interface by contacting E193 (in the 4jg3 crystal structure of Crc), whereas the authors only discuss its role in RNA binding. E193, instead is described by the authors to be relevant only in the Crc3-Crc4 interface. A more balanced and careful discussion of these interfaces is therefore suggested.

6) Subsection “A specialised and recurring RNA conformation in Hfq-mediated regulation”, last paragraph: The molecular determinants for A and R specificity are among the more central issues for this field, but this is not covered extensively. H-bonds consistent with a specificity pocket are observed and noted (Figure 4), but these do not necessarily confer specificity. Have the amino acids interacting with the N6 amine and N1 been mutated and tested previously and is this why the authors have not made these mutations? If so, the results should be discussed. That said, it would seem reasonable to test the effect of these mutants in the context of this more biologically relevant RNA sequence.

7) Discussion, third paragraph: It would be best if the stepwise assembly process of the higher-order complex were tested experimentally through order of additional experiments, SEC-mals or similarly informative biophysical experiments given the emphasis on the 2:4:2 complex given within the paper. This gets back to point #1, above. Otherwise, the authors need to change the speculative language in the text.

Reviewer #2:

This work presents exciting high resolution cryo EM structures (and some functional validation experiments) showing how the Crc 'helper' protein recognizes a composite Hfq-RNA surface and thus can act as a cofactor of Hfq-mediated RNA regulation in bacteria. The structure represents an important advance in our molecular understanding of Hfq function and demonstrates that it should be feasible to analyze similar Hfq complexes by cryo EM in the future. I'd enthusiastically support publication of the work in eLife after the overall description of the structure is improved to be more amenable to more generally interested readers.

The structure is not very clearly described and structure description is difficult to follow even for readers who have previously seen similar structures. A more streamlined description of the overall structure going from known interactions (Hfq-RNA) to the novel and exciting features (role of Crc in complex assembly), and a precise assignment of interface names (without swapping between symmetry-related interfaces) would help the general reader to better appreciate the principles of complex assembly and RNA recognition in the present case. Otherwise the most exciting insight from this work, i.e. how Crc improves RNA recognition and stabilizes the Hfq-RNA interaction gets lost too much in structural details (see point 2).

The Introduction could profit from some clarifications to help general readers understand and evaluate the significance of the present work (see point 1).

Some of the references to data by Sonnleitner et al., 2018, are rather vague and corresponding statements remain obscure without detailed prior knowledge of the work by Sonnleitner (see point 3).

1) Introduction

1.1) Clarify early in the general introduction of Hfq that Hfq controls mRNA translation not only indirectly (via small RNAs) but also directly (i.e. independently of small RNAs) by binding A-rich sequences at/near the mRNA translation initiation sites. Later on in the Introduction, mention that amiE RNA is such a case where Hfq's distal side binds A-rich sequences. This would also be a good occasion to mention that A-rich RNA binding to Hfq is principally understood in structural terms (Link et al.), but that the role of Crc in this context remained obscure and is the topic of the present study.

1.2) Clarify in the Introduction that Crc is a rather specialized protein that is not generally found in all bacteria but rather in a limited subset including Pseudomonas but excluding e.g. Enterobacteriaceae such as Escherichia coli. Such a statement does not limit the scope of the paper but helps the reader understand why the presented structure is so useful also for the design of selective drugs/inhibitors.

1.3) Mention in the Introduction that Crc is a very interesting member of the EEP protein family (Exonuclease/Endonuclease/Phosphatase) although key active site residues in the active site cleft are mutated. Information of the orientation of the active site cleft in the complex may be helpful in overall structure description (facing outward, laterally). Although Crc is catalytically inactive, I wonder whether it might bind phosphorylated sugars or metabolites in its active site cleft and hence helps to sense available carbon sources. Is anything known in this direction?

2) Structure description

2.1) Most importantly, I would suggest to improve and reorganize the description of the overall complex and of the individual interfaces. The authors use a defined nucleotide sequence (5'-AAA-AAU-AAC-AAC-AAG-AAG-3', note it down at the beginning of the Results section), and they assign nucleotide numbers in their structure model. In contrast to Link et al., the RNA density is hence not averaged over the Hfq hexamer and the 5' and 3' ends are visible. Consequently, it should be possible to number Hfq monomers from one (nucleotides 1-3) to six (nucleotides 16-18) and use these Hfq monomer numbers in the description of the various interfaces.

2.2) It seems more 'intuitive' to begin complex description with the Hfq-RNA complex and to point out what is novel/unique/different as compared to previous such complexes published by Link et al., 2009, or also Horstmann et al., 2012. In the following, all of the Crc interactions (i.e. with Crc1, Crc2 and Crc3) can then be described with only one of the Hfq-RNA complexes (since all of the interactions with the second Hfq ring are symmetry-related).

2.3) In a second step, one then could describe how the twofold symmetric Crc1-Crc2 dimer recognizes and binds over the Hfq-RNA interface, how the Hfq-RNA complex stabilizes the Crc1-Crc2 dimer and why the 2:2:2 complex forms cooperatively in an all-or-nothing fashion. Here one can then also describe and validate the Crc1-Crc2 interface in comparison to existing crystal packing interactions in Milojevic et al., 2013, (but also previously in Wei et al., 2012).

2.4) In a third step, one would then describe how Crc3 recognizes the Hfq-RNA complex, maybe in comparison to Crc1 and Crc2. Does Crc3 interact at all with Crc1 or Crc2? Finally, one could compare the Crc3-Crc4 interface to the Crc1-Crc2 interface and describe how the two Crc dimers are oriented with respect to each other in the full complex. As the authors point out, the Crc3-Crc4 interface is somewhat less important than the Crc1-Crc2 interface, because Crc3 and Crc4 can be added in a stepwise fashion.

3) Structure validation

3.1) Regarding structure validation, the authors mainly test for the different Crc dimer interfaces and only using duplicates in the translational repression reporter assay. It might be worthwhile to solidify these results and also test E142A and E193A mutants in order to see whether charge compensation really is the correct interpretation for the success of the rescue experiments. It also seems from the Crc crystal structure that E193 participates as well in the Crc1-Crc2 dimer interface. Maybe tone done a bit the exclusive importance of E193 for the Crc3-Crc4 interface. Moreover, since the authors have access to size exclusion chromatography and MALS instrumentation, they could for example test as well to which degree changes in the RNA sequence affect recognition by Crc and complex formation. Although not strictly required, these experiments clearly would strengthen the paper.

3.2) On various occasions, the authors rather vaguely point to previous results obtained by Sonnleitner et al., 2018, in support of the present cryo-EM structure. It would be good to also cite numeric values (e.g. subsection “An ensemble of Hfq/Crc/amiE_6ARNRNA assemblies”, first paragraph, ~220 kDa?) or define more precisely which mutants, experiments or data they have in mind when citing this paper (e.g. subsection “Function, origins and validation of subunit cooperativity in the 2:2:2 complex”, third paragraph).

3.3) Finally, in the Discussion, the authors might want to stress a bit more that this is the first time that structural information was obtained on how a 'helper' protein of Hfq can assist and specify the role of Hfq in direct translational repression. They also might want to explain a bit more extensively and speculate on what is the added value of Crc in the complex as compared to a simple recognition of the A-rich sequence by Hfq. In the second paragraph of the Discussion, it would again be important to be more precise and distinguish between thermodynamic and kinetic stability – otherwise this passage of the text is quite confusing.

Reviewer #3:

This study reveals for the first time the molecular mechanism by which two major regulatory proteins Hfq and Crc cooperate to regulate the translation of mRNAs in Pseudomonas aeruginosa. High-resolution structures of complexes formed by Hfq, Crc and the operator regulatory region of amiE mRNA were solved using cryo-EM. Based on these structures, specific mutations were introduced into Crc to analyze their ability to form RNP and to regulate amiE translation. Taken together, the authors visualized how Crc and Hfq assemble around a specific binding site on amiE mRNA to regulate translation initiation. This work is incredible and illustrates the importance of RNP assembly in the regulation of gene expression in bacteria. This work beautifully showed how a strong RNA binder (Hfq) is able to recruit proteins that are not able solely to bind RNA (Crc) in order to expand the regulatory networks. This work is incredible and opens up new ideas, i.e. to search for novel translational regulators, but also should pave the way to identify novel strategies interfering with the metabolism of pathogenic bacteria. This is an important paper, which also reconciles previous published works on Crc function.

eLife. 2019 Feb 13;8:e43158. doi: 10.7554/eLife.43158.029

Author response

Reviewer #1:

This is a fascinating manuscript that not only provides new insights into the molecular mechanism of Hfq function and its interplay with cofactors, but also contains lessons for understanding regulatory RNA-protein complexes in general. The structures themselves are intriguing and they provide a trove of information on molecular recognition strategies within RNA-protein complexes. For example, Figure 2 is particularly interesting as no single type of interaction is completely new, but the way they are used on concert is extremely informative. The one weakness of the paper (in addition to a relatively poor presentation of the different types of complexes) is the weak supporting data on the protein-protein interactions. Attention to the points below would strengthen the paper and enhance its impact.

1)Subsection “Protein synthesis, purification and complex formation”: More experimental details about complex assembly are useful but also important for mechanistic understanding. Was it necessary to incubate the RNA and the Hfq hexamer first before adding Crc? What order of addition experiments were done to arrive at this procedure? How did the authors determine the stoichiometry for Crc? Combining the RNA/Hfq/Crc in a 1:1:1 ratio is hard to understand given the ultimate stoichiometries observed in the particles – so why didn't the authors go back and try again? This is important, as it is germane to the biological relevance described later in the paper. Given the distribution of particle stoichiometries (Figure 1), why were other assembly conditions not tried and if they were, why were they not used? Some of this is discussed in Sonnleitner, and specific results from that paper could be cited more clearly in support of the present manuscript.

We have not exhaustively explored the order of addition to optimise the reconstitution, but we reasoned that the Hfq/RNA interaction is the stronger interaction and would nucleate the assembly. The conditions and search for optimal conditions were described in Sonnleitner et al., 2018, and we have added a comment to the revised text (first paragraph: Materials and methods). We also mentioned that SEC-MALS (Sonnleitner et al., 2018) have excluded a simple 1:1:1 assembly (subsection “An ensemble of Hfq/Crc/amiE_6ARNRNA assemblies”).

2) Subsection “An ensemble of Hfq/Crc/amiE_6ARNRNA assemblies”, first paragraph: Mild crosslinking conditions were purportedly used. Were others tried? The end of the paragraph says crosslinking does not affect structure. So why was it used? What did it help?

We used the bifunctional crosslinker BS3 to prepare specimens for all high-resolution datasets to maintain the complexes during grid preparations. To ensure that the observed assemblies were not artefacts from the crosslinking procedure, we subsequently collected a smaller data set without crosslinker and found that the structure remained unchanged within the resolution limits of the data for the models being compared. This is now explicitly mentioned in the subsection “An ensemble of Hfq/Crc/amiE_6ARNRNA assemblies”, and the results are presented in Supplementary file 1, Table S1. Given that no crosslinking artefacts were found, we did not further use/optimise the non-crosslinked datasets.

3) Subsection “An ensemble of Hfq/Crc/amiE_6ARNRNA assemblies”, last paragraph: In the first part of the paper, there is too little focus on the overall organization and features of the 2:2:2 complex and far too much focus on the organization of complexes in the different stoichiometries – it distracts from the basic elements that are common to all three and which are likely to be of central importance. For example, the discussion of symmetry is interesting, but not central and will not help with general understanding of function by most readers. Instead, in this part of the paper, it would be good to focus on overall organization of the 2:2:2 given that it is later suggested to be the most relevant one. In the opinion of this reviewer, the first part of the paper should not have figures and text with such an emphasis on alternative stoichiometries.

We have re-organised the text to focus first on the 2:2:2 complex, and then turn to the other complexes.

4) The pull-down experimental data shown in Figure 2—figure supplement 1 are qualitative and not well explained. It's not clear to the reader how the raw data demonstrate complex formation or specific interfaces unambiguously, at least as shown. Given the affinities known for these proteins, using EMSA or size/exclusion+MALS as experimental techniques on mutants would significantly strengthen this paper. That said, a clear and compelling explanation of why the existing data are sufficient might obviate the need for these experiments.

The IP data in Figure 2—figure supplement 1A are qualitative but suggest that there is a good agreement between the in vivo data and the in vitro results, indicating that the purified components are sufficient to recapitulate the complexity of the in vivo process. The method for the pull down is described in Sonnleitner et al., 2018, and we have changed the text to explain the data more clearly (subsection “Function, origins and validation of subunit cooperativity in the 2:2:2 complex”, second paragraph).

5) Subsection “Function, origins and validation of subunit cooperativity in the 2:2:2 complex”, third paragraph. The authors state that mutant Crc_R140E did not co-IP with Hfq – but the data in the Figure 2—figure supplement 1B look just like WT Crc, so this seems to undermine the authors' point, unless there is some type of misunderstanding. Another issue: it appears from the structure that R140 also participates in the Crc1-Crc2 interface by contacting E193 (in the 4jg3 crystal structure of Crc), whereas the authors only discuss its role in RNA binding. E193, instead is described by the authors to be relevant only in the Crc3-Crc4 interface. A more balanced and careful discussion of these interfaces is therefore suggested.

The data in Figure 2—figure supplement 1B show the input levels for the proteins used in the pulldowns for Figure 2—figure supplement 1A. In the 4jg3 crystal structure R140 indeed participates in the Crc1-Crc2 dimer interface, however, in our maps the density for R140 clearly shows a rotamer shift pulling the R140 side chain away from E193 and towards A3 of amiE in the Entry/Exit site. This has now been addressed in the revised text. E193 does participate in the dimer interface of Crc3-Crc4 via contacts with R230 and R233 (Figure 4A and 4C), as discussed in the text.

6) Subsection “A specialised and recurring RNA conformation in Hfq-mediated regulation”, last paragraph: The molecular determinants for A and R specificity are among the more central issues for this field, but this is not covered extensively. H-bonds consistent with a specificity pocket are observed and noted (Figure 4), but these do not necessarily confer specificity. Have the amino acids interacting with the N6 amine and N1 been mutated and tested previously and is this why the authors have not made these mutations? If so, the results should be discussed. That said, it would seem reasonable to test the effect of these mutants in the context of this more biologically relevant RNA sequence.

The network of hydrogen bonding interactions with the A and R in the A-R-N pattern in Pseudomonas aeruginosa Hfq are consistent with the interactions observed by Link et al., 2009, for the E. coli Hfq with polyA₁₈. The interactions do specify sequence preferences in a way that is consistent with binding affinities and rationalisation from the structure. For instance, the interactions with N1 (by the carbonyl of Q33) and N7 (by the amide of Q33) specify A as they are not compatible with a G, which form a repulsive contact of the O6 and peptide carbonyl. These interactions cannot be tested by mutagenesis because they do not involve side chain changes. Other changes in Hfq cannot be tested easily in vivo because of the pleotropic effect on many other RNA processes mediated by Hfq. We have changed the text to discuss the basis for the specificity.

We would also like to mention that our colleagues Jiri Sponer and Miroslav Krepl have found that the A in the A-R-N can switch between syn and anti-conformations in molecular dynamics simulations, and that both conformations are only compatible with A.

7) Discussion, third paragraph: It would be best if the stepwise assembly process of the higher-order complex were tested experimentally through order of additional experiments, SEC-mals or similarly informative biophysical experiments given the emphasis on the 2:4:2 complex given within the paper. This gets back to point #1, above. Otherwise, the authors need to change the speculative language in the text.

The proposed assembly process is inferred from the fact that the assemblies all share the same core 2:2:2 complex. Nevertheless, we have changed the text as suggested so that it is less speculative. The 2:4:2 complex is demonstrated to be important in vivo, as shown by the second site revertants (Figure 4B).

Reviewer #2:

This work presents exciting high resolution cryo EM structures (and some functional validation experiments) showing how the Crc 'helper' protein recognizes a composite Hfq-RNA surface and thus can act as a cofactor of Hfq-mediated RNA regulation in bacteria. The structure represents an important advance in our molecular understanding of Hfq function and demonstrates that it should be feasible to analyze similar Hfq complexes by cryo EM in the future. I'd enthusiastically support publication of the work in eLife after the overall description of the structure is improved to be more amenable to more generally interested readers.

The structure is not very clearly described and structure description is difficult to follow even for readers who have previously seen similar structures. A more streamlined description of the overall structure going from known interactions (Hfq-RNA) to the novel and exciting features (role of Crc in complex assembly), and a precise assignment of interface names (without swapping between symmetry-related interfaces) would help the general reader to better appreciate the principles of complex assembly and RNA recognition in the present case. Otherwise the most exciting insight from this work, i.e. how Crc improves RNA recognition and stabilizes the Hfq-RNA interaction gets lost too much in structural details (see point 2).

The Introduction could profit from some clarifications to help general readers understand and evaluate the significance of the present work (see point 1).

Some of the references to data by Sonnleitner et al., 2018, are rather vague and corresponding statements remain obscure without detailed prior knowledge of the work by Sonnleitner (see point 3).

1) Introduction

1.1) Clarify early in the general introduction of Hfq that Hfq controls mRNA translation not only indirectly (via small RNAs) but also directly (i.e. independently of small RNAs) by binding A-rich sequences at/near the mRNA translation initiation sites. Later on in the Introduction, mention that amiE RNA is such a case where Hfq's distal side binds A-rich sequences. This would also be a good occasion to mention that A-rich RNA binding to Hfq is principally understood in structural terms (Link et al.), but that the role of Crc in this context remained obscure and is the topic of the present study.

We have changed the “Introduction” as suggested.

1.2) Clarify in the Introduction that Crc is a rather specialized protein that is not generally found in all bacteria but rather in a limited subset including Pseudomonas but excluding e.g. Enterobacteriaceae such as Escherichia coli. Such a statement does not limit the scope of the paper but helps the reader understand why the presented structure is so useful also for the design of selective drugs/inhibitors.

We have made the suggested changes.

1.3) Mention in the Introduction that Crc is a very interesting member of the EEP protein family (Exonuclease/Endonuclease/Phosphatase) although key active site residues in the active site cleft are mutated. Information of the orientation of the active site cleft in the complex may be helpful in overall structure description (facing outward, laterally). Although Crc is catalytically inactive, I wonder whether it might bind phosphorylated sugars or metabolites in its active site cleft and hence helps to sense available carbon sources. Is anything known in this direction?

We have included the suggested changes in the text. The suggestion that Crc might bind metabolites is an intriguing concept, but we are not aware of any evidence for this.

2) Structure description

2.1) Most importantly, I would suggest to improve and reorganize the description of the overall complex and of the individual interfaces. The authors use a defined nucleotide sequence (5'-AAA-AAU-AAC-AAC-AAG-AAG-3', note it down at the beginning of the Results section), and they assign nucleotide numbers in their structure model. In contrast to Link et al., the RNA density is hence not averaged over the Hfq hexamer and the 5' and 3' ends are visible. Consequently, it should be possible to number Hfq monomers from one (nucleotides 1-3) to six (nucleotides 16-18) and use these Hfq monomer numbers in the description of the various interfaces.

The proposed assignment numbers to Hfq monomers and the corresponding A-R-N nucleotides might help to describe each ARN-interface independently, but we feel that this would be unnecessary because they are all very similar.

2.2) It seems more 'intuitive' to begin complex description with the Hfq-RNA complex and to point out what is novel/unique/different as compared to previous such complexes published by Link et al., 2009, or also Horstmann et al., 2012. In the following, all of the Crc interactions (i.e. with Crc1, Crc2 and Crc3) can then be described with only one of the Hfq-RNA complexes (since all of the interactions with the second Hfq ring are symmetry-related).

We have reorganised the text to begin the description of the Hfq-RNA interactions and included a section in the “Results” and in the “Discussion” to compare our structures to that of Link et al. and Horstman et al. structures (and reference to Horstmann et al., 2012). A supplementary figure to show this has been added as well (Figure 2—figure supplement 1).

2.3) In a second step, one then could describe how the twofold symmetric Crc1-Crc2 dimer recognizes and binds over the Hfq-RNA interface, how the Hfq-RNA complex stabilizes the Crc1-Crc2 dimer and why the 2:2:2 complex forms cooperatively in an all-or-nothing fashion. Here one can then also describe and validate the Crc1-Crc2 interface in comparison to existing crystal packing interactions in Milojevic et al., 2013, (but also previously in Wei et al., 2012).

We have made the suggested change to the text and included the reference Wei et al., 2012.

2.4) In a third step, one would then describe how Crc3 recognizes the Hfq-RNA complex, maybe in comparison to Crc1 and Crc2. Does Crc3 interact at all with Crc1 or Crc2? Finally, one could compare the Crc3-Crc4 interface to the Crc1-Crc2 interface and describe how the two Crc dimers are oriented with respect to each other in the full complex. As the authors point out, the Crc3-Crc4 interface is somewhat less important than the Crc1-Crc2 interface, because Crc3 and Crc4 can be added in a stepwise fashion.

We have made the suggested change to the text.

3) Structure validation

3.1) Regarding structure validation, the authors mainly test for the different Crc dimer interfaces and only using duplicates in the translational repression reporter assay. It might be worthwhile to solidify these results and also test E142A and E193A mutants in order to see whether charge compensation really is the correct interpretation for the success of the rescue experiments. It also seems from the Crc crystal structure that E193 participates as well in the Crc1-Crc2 dimer interface. Maybe tone done a bit the exclusive importance of E193 for the Crc3-Crc4 interface. Moreover, since the authors have access to size exclusion chromatography and MALS instrumentation, they could for example test as well to which degree changes in the RNA sequence affect recognition by Crc and complex formation. Although not strictly required, these experiments clearly would strengthen the paper.

As suggested, we have introduced the E142A exchange in the R229/230E background and the E193A exchange in the R230E background. The triple substitution Crc_{E142A,R229E,R230E} has a similar in vivo activity as the Crc_{E142R,R229E,R230E} variant.

For the Crc_E193A,R230E variant, the repressive activity was also comparable to that of the Crc_{E193R, R230} mutant protein.

These results suggest that the rescue of the deleterious mutations does not require charge compensation, but only removing any potential charge repulsion. We have mentioned this in the revised text.

In the 4jg3 crystal structure R140 indeed participates in the Crc1-Crc2 dimer interface, however, in our maps the density for R140 clearly shows a rotamer shift pulling away the R140 sidechain from E193 and towards A3 from amiE in the Entry/Exit site. This has now been mentioned in the revised text. E193 does participate in the dimer interface of Crc3-Crc4 via contacts with R230 and R233 (Figure 4A and 4C), as discussed in the text.

3.2) On various occasions, the authors rather vaguely point to previous results obtained by Sonnleitner et al., 2018, in support of the present cryo-EM structure. It would be good to also cite numeric values (e.g. subsection “An ensemble of Hfq/Crc/amiE_6ARNRNA assemblies”, first paragraph, ~220 kDa?) or define more precisely which mutants, experiments or data they have in mind when citing this paper (e.g. subsection “Function, origins and validation of subunit cooperativity in the 2:2:2 complex”, third paragraph).

We have clarified the text.

3.3) Finally, in the Discussion, the authors might want to stress a bit more that this is the first time that structural information was obtained on how a 'helper' protein of Hfq can assist and specify the role of Hfq in direct translational repression. They also might want to explain a bit more extensively and speculate on what is the added value of Crc in the complex as compared to a simple recognition of the A-rich sequence by Hfq. In the second paragraph of the Discussion, it would again be important to be more precise and distinguish between thermodynamic and kinetic stability – otherwise this passage of the text is quite confusing.

We have changed the text as suggested.

Associated Data

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

Data Citations

Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Blasi, Ben F Luisi. 2019. Hfq/Crc/RNA cpmplex 2:4:2 ratio. Protein Data Bank in Europe. 6o1m
Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Blasi, Ben F Luisi. 2019. Hfq/Crc/RNA complex 2:2:2 ratio. Electron Microscopy Data Bank. EMD-0590
Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Blasi, Ben F Luisi. 2019. Hfq/Crc/RNA complex 2:3:2 ratio. Electron Microscopy Data Bank. EMD-0591
Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Blasi, Ben F Luisi. 2019. Hfq/Crc/RNA complex 2:4:2 ratio. Electron Microscopy Data Bank. EMD-0592
Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Bläsi, Ben F Luisi. 2019. Hfq/Crc/RNA complex 2:2:2 ratio. Protein Data Bank in Europe. 6o1k
Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Bläsi. 2019. Hfq/Crc/RNA complex 2:3:2 ratio. Protein Data Bank in Europe. 6o1l

Supplementary Materials

Supplementary file 1. Resolution of maps and angular distributions of 2D images.

elife-43158-supp1.docx^{(39.6KB, docx)}

DOI: 10.7554/eLife.43158.013

Transparent reporting form

elife-43158-transrepform.docx^{(246.6KB, docx)}

DOI: 10.7554/eLife.43158.014

Data Availability Statement

CryoEM data have been deposited in the Electron Microscopy Data Bank.

The following datasets were generated:

Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Blasi, Ben F Luisi. 2019. Hfq/Crc/RNA cpmplex 2:4:2 ratio. Protein Data Bank in Europe. 6o1m

Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Blasi, Ben F Luisi. 2019. Hfq/Crc/RNA complex 2:2:2 ratio. Electron Microscopy Data Bank. EMD-0590

Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Blasi, Ben F Luisi. 2019. Hfq/Crc/RNA complex 2:3:2 ratio. Electron Microscopy Data Bank. EMD-0591

Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Blasi, Ben F Luisi. 2019. Hfq/Crc/RNA complex 2:4:2 ratio. Electron Microscopy Data Bank. EMD-0592

Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Bläsi, Ben F Luisi. 2019. Hfq/Crc/RNA complex 2:2:2 ratio. Protein Data Bank in Europe. 6o1k

Tom Dendooven, Xue Yuan Pei, Elisabeth Sonnleitner, Shaoxia Chen, Udo Bläsi. 2019. Hfq/Crc/RNA complex 2:3:2 ratio. Protein Data Bank in Europe. 6o1l

[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] Bandyra KJ, Sinha D, Syrjanen J, Luisi BF, De Lay NR. The ribonuclease polynucleotide phosphorylase can interact with small regulatory RNAs in both protective and degradative modes. RNA. 2016;22:360–372. doi: 10.1261/rna.052886.115. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib3] Chen S, McMullan G, Faruqi AR, Murshudov GN, Short JM, Scheres SH, Henderson R. High-resolution noise substitution to measure overfitting and validate resolution in 3D structure determination by single particle electron cryomicroscopy. Ultramicroscopy. 2013;135:24–35. doi: 10.1016/j.ultramic.2013.06.004. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib4] Clamens T, Rosay T, Crépin A, Grandjean T, Kentache T, Hardouin J, Bortolotti P, Neidig A, Mooij M, Hillion M, Vieillard J, Cosette P, Overhage J, O'Gara F, Bouffartigues E, Dufour A, Chevalier S, Guery B, Cornelis P, Feuilloley MG, Lesouhaitier O. The aliphatic amidase AmiE is involved in regulation of pseudomonas aeruginosa virulence. Scientific Reports. 2017;7:41178. doi: 10.1038/srep41178. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib5] Croll TI. ISOLDE : a physically realistic environment for model building into low-resolution electron-density maps. Acta Crystallographica Section D Structural Biology. 2018;74:519–530. doi: 10.1107/S2059798318002425. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib6] 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]

[bib7] Fender A, Elf J, Hampel K, Zimmermann B, Wagner EGH. RNAs actively cycle on the Sm-like protein hfq. Genes & Development. 2010;24:2621–2626. doi: 10.1101/gad.591310. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib8] Fernández L, Breidenstein EB, Taylor PK, Bains M, de la Fuente-Núñez C, Fang Y, Foster LJ, Hancock RE. Interconnection of post-transcriptional regulation: the RNA-binding protein hfq is a novel target of the lon protease in Pseudomonas aeruginosa. Scientific Reports. 2016;6:26811. doi: 10.1038/srep26811. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib9] Heitzinger DA. University of Vienna; 2016. Impact of Hfq on Antibiotic Susceptibility of Pseudomonas Aeruginosa. Master Thesis. [Google Scholar]

[bib10] Holloway BW, Krishnapillai V, Morgan AF. Chromosomal genetics of pseudomonas. Microbiological Reviews. 1979;43:73–102. doi: 10.1128/mr.43.1.73-102.1979. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib11] Horstmann N, Orans J, Valentin-Hansen P, Shelburne SA, Brennan RG. Structural mechanism of staphylococcus aureus Hfq binding to an RNA A-tract. Nucleic Acids Research. 2012;40:11023–11035. doi: 10.1093/nar/gks809. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib12] Huang J, Sonnleitner E, Ren B, Xu Y, Haas D. Catabolite repression control of pyocyanin biosynthesis at an intersection of primary and secondary metabolism in pseudomonas aeruginosa. Applied and Environmental Microbiology. 2012;78:5016–5020. doi: 10.1128/AEM.00026-12. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib13] Ikeda Y, Yagi M, Morita T, Aiba H. Hfq binding at RhlB-recognition region of RNase E is crucial for the rapid degradation of target mRNAs mediated by sRNAs in Escherichia coli. Molecular Microbiology. 2011;79:419–432. doi: 10.1111/j.1365-2958.2010.07454.x. [DOI] [PubMed] [Google Scholar]

[bib14] Jakobi AJ, Wilmanns M, Sachse C. Model-based local density sharpening of cryo-EM maps. eLife. 2017;6:e27131. doi: 10.7554/eLife.27131. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib15] Kambara TK, Ramsey KM, Dove SL. Pervasive targeting of nascent transcripts by hfq. Cell Reports. 2018;23:1543–1552. doi: 10.1016/j.celrep.2018.03.134. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib16] Linares JF, Moreno R, Fajardo A, Martínez-Solano L, Escalante R, Rojo F, Martínez JL. The global regulator crc modulates metabolism, susceptibility to antibiotics and virulence in Pseudomonas aeruginosa. Environmental Microbiology. 2010;12:3196–3212. doi: 10.1111/j.1462-2920.2010.02292.x. [DOI] [PubMed] [Google Scholar]

[bib17] Link TM, Valentin-Hansen P, Brennan RG. Structure of Escherichia coli hfq bound to polyriboadenylate RNA. PNAS. 2009;106:19292–19297. doi: 10.1073/pnas.0908744106. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib18] Lu P, Wang Y, Zhang Y, Hu Y, Thompson KM, Chen S. RpoS-dependent sRNA RgsA regulates fis and AcpP in Pseudomonas aeruginosa. Molecular Microbiology. 2016;102:244–259. doi: 10.1111/mmi.13458. [DOI] [PubMed] [Google Scholar]

[bib19] Miller JH. Experiments in Molecular Genetics. New York: Cold Spring Harbor Press; 1972. [Google Scholar]

[bib20] Milojevic T, Grishkovskaya I, Sonnleitner E, Djinovic-Carugo K, Bläsi U. The Pseudomonas aeruginosa catabolite repression control protein Crc is devoid of RNA binding activity. PLOS ONE. 2013;8:e64609. doi: 10.1371/journal.pone.0064609. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib21] Mohanty BK, Maples VF, Kushner SR. The Sm-like protein hfq regulates polyadenylation dependent mRNA decay in Escherichia coli. Molecular Microbiology. 2004;54:905–920. doi: 10.1111/j.1365-2958.2004.04337.x. [DOI] [PubMed] [Google Scholar]

[bib22] Moreno R, Hernández-Arranz S, La Rosa R, Yuste L, Madhushani A, Shingler V, Rojo F. The crc and hfq proteins of Pseudomonas putida cooperate in catabolite repression and formation of ribonucleic acid complexes with specific target motifs. Environmental Microbiology. 2015;17:105–118. doi: 10.1111/1462-2920.12499. [DOI] [PubMed] [Google Scholar]

[bib23] 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]

[bib24] O'Toole GA, Gibbs KA, Hager PW, Phibbs PV, Kolter R. The global carbon metabolism regulator crc is a component of a signal transduction pathway required for biofilm development by pseudomonas aeruginosa. Journal of Bacteriology. 2000;182:425–431. doi: 10.1128/JB.182.2.425-431.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib25] Panja S, Schu DJ, Woodson SA. Conserved arginines on the rim of hfq catalyze base pair formation and exchange. Nucleic Acids Research. 2013;41:7536–7546. doi: 10.1093/nar/gkt521. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib26] Pantelic RS, Meyer JC, Kaiser U, Baumeister W, Plitzko JM. Graphene oxide: a substrate for optimizing preparations of frozen-hydrated samples. Journal of Structural Biology. 2010;170:152–156. doi: 10.1016/j.jsb.2009.12.020. [DOI] [PubMed] [Google Scholar]

[bib27] Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, Ferrin TE. UCSF chimera--a visualization system for exploratory research and analysis. Journal of Computational Chemistry. 2004;25:1605–1612. doi: 10.1002/jcc.20084. [DOI] [PubMed] [Google Scholar]

[bib28] 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]

[bib29] Pusic P, Tata M, Wolfinger MT, Sonnleitner E, Häussler S, Bläsi U. Cross-regulation by CrcZ RNA controls anoxic biofilm formation in pseudomonas aeruginosa. Scientific Reports. 2016;6:39621. doi: 10.1038/srep39621. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib30] Ramírez-aportela E, Vilas JL, Melero R, Conesa P. Automatic local resolution-based sharpening of cryo-EM maps. bioRxiv. 2018 doi: 10.1101/433284. [DOI] [PMC free article] [PubMed]

[bib31] Resch A, Većerek B, Palavra K, Bläsi U. Requirement of the CsdA DEAD-box helicase for low temperature riboregulation of rpoS mRNA. RNA Biology. 2010;7:796–802. doi: 10.4161/rna.7.6.13768. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib32] Rojo F. Carbon catabolite repression in Pseudomonas : optimizing metabolic versatility and interactions with the environment. FEMS Microbiology Reviews. 2010;34:658–684. doi: 10.1111/j.1574-6976.2010.00218.x. [DOI] [PubMed] [Google Scholar]

[bib33] Salvail H, Caron MP, Bélanger J, Massé E. Antagonistic functions between the RNA chaperone hfq and an sRNA regulate sensitivity to the antibiotic colicin. The EMBO Journal. 2013;32:2764–2778. doi: 10.1038/emboj.2013.205. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib34] Santiago-Frangos A, Kavita K, Schu DJ, Gottesman S, Woodson SA. C-terminal domain of the RNA chaperone hfq drives sRNA competition and release of target RNA. PNAS. 2016;113:E6089–E6096. doi: 10.1073/pnas.1613053113. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib35] Santiago-Frangos A, Jeliazkov JR, Gray JJ, Woodson SA. Acidic C-terminal domains autoregulate the RNA chaperone hfq. eLife. 2017;6:e27049. doi: 10.7554/eLife.27049. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib36] Sauer E, Schmidt S, Weichenrieder O. Small RNA binding to the lateral surface of hfq hexamers and structural rearrangements upon mRNA target recognition. PNAS. 2012;109:9396–9401. doi: 10.1073/pnas.1202521109. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib37] Scheres SH. A bayesian view on cryo-EM structure determination. Journal of Molecular Biology. 2012;415:406–418. doi: 10.1016/j.jmb.2011.11.010. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib38] Scheres SH, Chen S. Prevention of overfitting in cryo-EM structure determination. Nature Methods. 2012;9:853–854. doi: 10.1038/nmeth.2115. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib39] Schulz EC, Seiler M, Zuliani C, Voigt F, Rybin V, Pogenberg V, Mücke N, Wilmanns M, Gibson TJ, Barabas O. Intermolecular base stacking mediates RNA-RNA interaction in a crystal structure of the RNA chaperone hfq. Scientific Reports. 2017;7:9903. doi: 10.1038/s41598-017-10085-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib40] Schumacher MA, Pearson RF, Møller T, Valentin-Hansen P, Brennan RG. Structures of the pleiotropic translational regulator hfq and an Hfq-RNA complex: a bacterial Sm-like protein. The EMBO Journal. 2002;21:3546–3556. doi: 10.1093/emboj/cdf322. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib41] Sonnleitner E, Hagens S, Rosenau F, Wilhelm S, Habel A, Jäger KE, Bläsi U. Reduced virulence of a hfq mutant of pseudomonas aeruginosa O1. Microbial Pathogenesis. 2003;35:217–228. doi: 10.1016/S0882-4010(03)00149-9. [DOI] [PubMed] [Google Scholar]

[bib42] Sonnleitner E, Schuster M, Sorger-Domenigg T, Greenberg EP, Bläsi U. Hfq-dependent alterations of the transcriptome profile and effects on quorum sensing in pseudomonas aeruginosa. Molecular Microbiology. 2006;59:1542–1558. doi: 10.1111/j.1365-2958.2006.05032.x. [DOI] [PubMed] [Google Scholar]

[bib43] Sonnleitner E, Abdou L, Haas D. Small RNA as global regulator of carbon catabolite repression in pseudomonas aeruginosa. PNAS. 2009;106:21866–21871. doi: 10.1073/pnas.0910308106. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib44] Sonnleitner E, Wulf A, Campagne S, Pei XY, Wolfinger MT, Forlani G, Prindl K, Abdou L, Resch A, Allain FH, Luisi BF, Urlaub H, Bläsi U. Interplay between the catabolite repression control protein crc, hfq and RNA in Hfq-dependent translational regulation in pseudomonas aeruginosa. Nucleic Acids Research. 2018;46:1470–1485. doi: 10.1093/nar/gkx1245. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib45] Sonnleitner E, Bläsi U. Regulation of hfq by the RNA CrcZ in Pseudomonas aeruginosa carbon catabolite repression. PLOS Genetics. 2014;10:e1004440. doi: 10.1371/journal.pgen.1004440. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib46] Sukhodolets MV, Garges S. Interaction of Escherichia coli RNA polymerase with the ribosomal protein S1 and the Sm-like ATPase Hfq. Biochemistry. 2003;42:8022–8034. doi: 10.1021/bi020638i. [DOI] [PubMed] [Google Scholar]

[bib47] Valentini M, García-Mauriño SM, Pérez-Martínez I, Santero E, Canosa I, Lapouge K. Hierarchical management of carbon sources is regulated similarly by the CbrA/B systems in pseudomonas aeruginosa and Pseudomonas putida. Microbiology. 2014;160:2243–2252. doi: 10.1099/mic.0.078873-0. [DOI] [PubMed] [Google Scholar]

[bib48] Van den Bossche A, Ceyssens PJ, De Smet J, Hendrix H, Bellon H, Leimer N, Wagemans J, Delattre AS, Cenens W, Aertsen A, Landuyt B, Minakhin L, Severinov K, Noben JP, Lavigne R. Systematic identification of hypothetical bacteriophage proteins targeting key protein complexes of pseudomonas aeruginosa. Journal of Proteome Research. 2014;13:4446–4456. doi: 10.1021/pr500796n. [DOI] [PubMed] [Google Scholar]

[bib49] Vecerek B, Moll I, Bläsi U. Translational autocontrol of the Escherichia coli hfq RNA chaperone gene. RNA. 2005;11:976–984. doi: 10.1261/rna.2360205. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib50] Vecerek B, Beich-Frandsen M, Resch A, Bläsi U. Translational activation of rpoS mRNA by the non-coding RNA DsrA and hfq does not require ribosome binding. Nucleic Acids Research. 2010;38:1284–1293. doi: 10.1093/nar/gkp1125. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib51] Vogel J, Luisi BF. Hfq and its constellation of RNA. Nature Reviews Microbiology. 2011;9:578–589. doi: 10.1038/nrmicro2615. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib52] Wagner EGH, Romby P. Small RNAs in bacteria and archaea: who they are, what they do, and how they do it. Advances in Genetics. 2015;90:133–208. doi: 10.1016/bs.adgen.2015.05.001. [DOI] [PubMed] [Google Scholar]

[bib53] Wei Y, Zhang H, Gao ZQ, Xu JH, Liu QS, Dong YH. Structure analysis of the global metabolic regulator crc from Pseudomonas aeruginosa. IUBMB Life. 2013;65:50–57. doi: 10.1002/iub.1103. [DOI] [PubMed] [Google Scholar]

[bib54] Williams CJ, Headd JJ, Moriarty NW, Prisant MG, Videau LL, Deis LN, Verma V, Keedy DA, Hintze BJ, Chen VB, Jain S, Lewis SM, Arendall WB, Snoeyink J, Adams PD, Lovell SC, Richardson JS, Richardson DC. MolProbity: more and better reference data for improved all-atom structure validation. Protein Science. 2018;27:293–315. doi: 10.1002/pro.3330. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib55] Worrall JA, Górna M, Crump NT, Phillips LG, Tuck AC, Price AJ, Bavro VN, Luisi BF. Reconstitution and analysis of the multienzyme Escherichia coli RNA degradosome. Journal of Molecular Biology. 2008;382:870–883. doi: 10.1016/j.jmb.2008.07.059. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib56] Yang N, Ding S, Chen F, Zhang X, Xia Y, Di H, Cao Q, Deng X, Wu M, Wong CC, Tian XX, Yang CG, Zhao J, Lan L. The crc protein participates in down-regulation of the lon gene to promote rhamnolipid production and rhl quorum sensing in pseudomonas aeruginosa. Molecular Microbiology. 2015;96:526–547. doi: 10.1111/mmi.12954. [DOI] [PubMed] [Google Scholar]

[bib57] Zhang L, Chiang WC, Gao Q, Givskov M, Tolker-Nielsen T, Yang L, Zhang G. The catabolite repression control protein Crc plays a role in the development of antimicrobial-tolerant subpopulations in Pseudomonas aeruginosa biofilms. Microbiology. 2012;158:3014–3019. doi: 10.1099/mic.0.061192-0. [DOI] [PubMed] [Google Scholar]

[bib58] Zhang L, Gao Q, Chen W, Qin H, Hengzhuang W, Chen Y, Yang L, Zhang G. Regulation of pqs quorum sensing via catabolite repression control in pseudomonas aeruginosa. Microbiology. 2013;159:1931–1936. doi: 10.1099/mic.0.066266-0. [DOI] [PubMed] [Google Scholar]

[bib59] 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]

[bib60] Zheng SQ, Palovcak E, Armache JP, 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]

[bib61] Zivanov J, Nakane T, Forsberg BO, Kimanius D, Hagen WJ, Lindahl E, Scheres SH. New tools for automated high-resolution cryo-EM structure determination in RELION-3. eLife. 2018;7:e42166. doi: 10.7554/eLife.42166. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Architectural principles for Hfq/Crc-mediated regulation of gene expression

Xue Yuan Pei

Tom Dendooven

Elisabeth Sonnleitner

Shaoxia Chen

Udo Bläsi

Ben F Luisi

Roles

Abstract

eLife digest

Introduction

Results

An ensemble of Hfq/Crc/amiE6ARNRNA assemblies

Figure 1. Reference free 2D classification and 3D classification of Hfq/Crc/RNA particles.

Figure 1—figure supplement 1. Resolution of maps and angular distributions of 2D images.

Figure 1—figure supplement 2. Raw micrograph image, fit of refined model into cryo-EM map, and correlation of experimental map and models.

A specialised and recurring RNA conformation in Hfq-mediated regulation

Figure 2. The 'A-R-N crown' in the Hfq/amiE6ARN RNA complex.

Figure 2—figure supplement 1. Interactions of RNA with the distal face of Hfq.

The quaternary structure of the core complex

Figure 3. Model of the 2:2:2 Hfq/Crc/RNA complex and validation of interactions.

Figure 3—figure supplement 1. Association of Hfq and Crc in the presence of RNA.

Function, origins and validation of subunit cooperativity in the 2:2:2 complex

Recruitment of additional Crc units to the core

Figure 4. Model of the 2:4:2 Hfq/Crc/RNA complex and validation of interactions.

Function, origins and validation of subunit cooperativity in the 2:4:2 complex

Video 1. A comparison of the Crc-Crc interface in the 2:2:2 Hfq:Crc:RNA complex (the core complex) and the second type of interface formed between the additional Crc components of the 2:4:2 complex.

Figure 5. Schematic pathway of carbon catabolite repression.

Discussion

Materials and methods

Key resources table.

Protein synthesis, purification and complex formation

CryoEM specimen preparation and data acquisition

Movie processing, single particle analysis, 3D reconstruction and refinement

Bacterial strains and plasmids

Construction of plasmids encoding Crc variant proteins for in vivo translational repression assay

In vivo translational repression of an amiE::lacZ reporter gene in the presence of Crc variants

Construction of plasmids employed for the production of selected Crc mutant proteins

Purification of Crc and Crc variants

In vitro co-IP studies

Western blot analyses

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

An ensemble of Hfq/Crc/amiE_6ARNRNA assemblies

Figure 2. The 'A-R-N crown' in the Hfq/amiE_6ARN RNA complex.