X-ray crystal structure of CqsR_p complexed with ethanolamine

To understand the mechanistic basis of ethanolamine recognition by CqsR, we determined the 1.55 Å resolution X-ray crystal structure of the CqsR periplasmic domain (CqsR_p) in complex with ethanolamine (Fig 2 and Table 1). To obtain this structure, CqsR_p crystals were soaked in cryoprotectant containing ethanolamine (S1 Table). However, it is important to note that subsequent LC-MS/MS analysis revealed that 103.3 μM CqsR_p is bound to 83.9 μM ethanolamine following overexpression in E. coli and purification (S1 Fig and S2 Table). Ultimately, we also obtained identical crystal structures without supplementing the cryoprotectant with ethanolamine and without adding it during crystallization.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Data collection and refinement statistics.

https://doi.org/10.1371/journal.ppat.1013447.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. CqsR_p-ethanolamine crystal structure.

(A) Left: The CqsR_p monomer and ethanolamine are rendered as a cyan cartoon and spheres, respectively. The schematic was produced using PyMOL (PyMOL Molecular Graphics System, Version 3.0 Schrödinger, LLC). Right: Expanded view of the area enclosed by the box in the left panel. F_o-F_c electron density corresponding to ethanolamine scaled to 3σ. Positive electron density is shown in green and negative density shown in red. F_o-F_c electron density was calculated prior to ligand and water building. The schematic was produced using Coot [34]. (B) Ethanolamine (ball-and-stick model) and its coordinating residues (cyan sticks) are shown. The schematic was produced using PyMOL (PyMOL Molecular Graphics System, Version 3.0 Schrödinger, LLC). (C) Schematic representation of CqsR_p-ethanolamine binding. CqsR_p and ethanolamine are depicted as blue and grey bonds, respectively. Hydrogen bonds are depicted as black dashed lines. Hydrophobic contacts are depicted as lines radiating from the semicircles and spheres. The schematic was produced using LigPlot+ [35].

https://doi.org/10.1371/journal.ppat.1013447.g002

Consistent with sequence analysis [30], structural comparison of CqsR_p to all structures in the PDB using Dali showed that it forms a double Cache domain (Fig 2A) [36]. Electron density corresponding to ethanolamine revealed that it binds to the core of the membrane distal domain (Fig 2A). The ethanolamine binding site consists of a hydrophilic platform and a hydrophobic lid (Fig 2B and 2C). More specifically, the acidic side chain of CqsR-D198 plays a critical role, serving to coordinate both the ethanolamine amine and hydroxyl. As both ethanolamine functional groups are chelated by Asp198, this enforces a higher energy gauche configuration for the ligand. CqsR residues Tyr169, Asp171, and Thr196 mediate additional contacts with ethanolamine. The aromatic hydrophobic side chains of residues Trp105, Phe123, Trp138, and Trp151 form a lid above ethanolamine, with the face of Trp151 prominently positioned to engage in hydrophobic contacts with the ethanolamine alkane carbon atoms.

X-ray crystal structure of apo CqsR_p (CqsR_p-D198N)

Because purified CqsR_p contains ethanolamine derived from the E. coli overexpression host, to obtain an apo, i.e., ligand-free, structure, we generated CqsR_p-D198N. Based on the CqsR_p-ethanolamine structure described above, we hypothesized that substituting Asp198 for Asn would disrupt Asp198 hydrogen bonding with ethanolamine without disrupting protein folding (Fig 2B and 2C). LC-MS/MS analysis of CqsR_p-D198N showed that it contains less than 2.1 ± 0.2 μM ethanolamine in a 103.3 μM sample of protein, and we then determined the X-ray crystal structure of apo CqsR_p-D198N to 2.50 Å resolution (Table 1, S1 and S2 Figs, and S3 Table). Since substituting negatively charged Asp198 for uncharged Asn resulted in a dramatic reduction in CqsR_p-ethanolamine binding affinity as determined by LC-MS/MS and the absence of ethanolamine in the crystal structure, we conclude that the Asp198 negative charge is critical for ethanolamine coordination.

Thermodynamic binding studies of CqsR to ethanolamine and ethanolamine analogs

We previously tested the common metabolites in the Biolog PM1–4 screens for their ability to bind CqsR_p as a function of its thermal unfolding [30]. Ethanolamine, L-alaninol, and serinol increased the CqsR_p melting temperature. These structurally similar 2-amino alcohols also acted as CqsR signaling agonists in vivo. To quantify their binding to CqsR, we used microscale thermophoresis (MST) (Fig 3). The dissociation constants (K_d) for CqsR_p binding to ethanolamine, L-alaninol, and serinol are 1.3 ± 0.1 μM (Fig 3A), 3.2 ± 0.1 μM (Fig 3B), and 15.9 ± 0.8 μM (Fig 3C), respectively. Importantly, ethanolamine, L-alaninol, and serinol did not bind CqsR_p-D198N (S3 Fig). In addition, we used MST to evaluate CqsR_p binding to additional diverse ethanolamine analogs (Fig 4). Most of these compounds exhibited no detectable binding to CqsR_p or otherwise bound very weakly (e.g., ethylenediamine K_d= 234.9 ± 28.9 μM, 2-amino-1-butanol K_d = 114.9 ± 40.3 μM, (S)-(+)-2-amino-1-butanol K_d = 115.0 ± 18.2 μM, and tris K_d= 3,112.2 ± 394.9 μM) compared to ethanolamine, L-alaninol, and serinol (Figs 4 and S4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. MST quantification of CqsR_p binding to ethanolamine, L-alaninol, and serinol.

(A) Ethanolamine was titrated between 500.0 μM and 15.3 nM with 100.0 nM CqsR_p. MST was performed in triplicate. (B) L-alaninol was titrated between 1.0 mM and 30.5 nM with 100.0 nM CqsR_p. MST was performed in quadruplicate. (C) Serinol was titrated between 5.0 mM and 152.6 nM with 100.0 nM CqsR_p. MST was performed in quadruplicate.

https://doi.org/10.1371/journal.ppat.1013447.g003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Dissociation constants derived from MST studies of CqsR_p binding to ethanolamine analogs.

Reported dissociation constants (K_d) for ethanolamine and compounds of similar chemical composition. Diagrams of each respective structure are shown. Nb indicates that no binding was detected.

https://doi.org/10.1371/journal.ppat.1013447.g004

X-ray crystal structures of CqsR in complex with the ethanolamine analogs L-alaninol or serinol

To understand how CqsR binds ethanolamine analogs, we determined the X-ray crystal structures of CqsR_p-L-alaninol and CqsR_p-serinol to 1.72 Å and 1.75 Å resolution, respectively (Table 1). In both cases, there is interpretable electron density corresponding to the entire ligand (Fig 5A and 5B). The L-alaninol and serinol atoms analogous to ethanolamine bound to CqsR_p in an identical fashion (Figs 2B, 2C and 5C–5F). More specifically, the amine and hydroxyl common to ethanolamine, L-alaninol, and serinol similarly hydrogen bond to the platform residue Asp198. Additional hydrogen bonds to the common amine and hydroxyl are mediated by Tyr169, Asp171, and Thr196 (Fig 5E and 5F). As in the CqsR_p-ethanolamine structure, aromatic hydrophobic side chains form a lid above the L-alaninol and serinol alkane carbon atoms. Relative to its position in the CqsR_p-ethanolamine structure, CqsR_p Trp138 shifts 0.7 Å and 0.6 Å to accommodate the L-alaninol methyl and serinol methoxy, respectively (Figs 2B, 5C, 5D and S5). Serinol binding also forms hydrogen bonds with Trp138 and Asp171 that are otherwise absent in the CqsR_p complexes with ethanolamine and L-alaninol (Fig 5D and 5F). The versatility of Trp138 is notable because it mediates hydrophobic contacts with ethanolamine (Fig 2B and 2C) and L-alaninol (Fig 5C and 5E) but forms a hydrogen bond with serinol (Fig 5D and 5F).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. CqsR_p-L-alaninol and CqsR_p-serinol binding sites.

(A and B) CqsR_p-L-alaninol and CqsR_p-serinol binding-site residues are depicted as balls and sticks. F_o-F_c maps scaled to 3σ with positive density shown in green and negative density shown in red. F_o-F_c electron density was calculated prior to ligand and water building. The schematic was produced using Coot [34]. (C and D) Isolated views of the respective ligand binding sites. CqsR residues that coordinate L-alaninol or serinol are shown as orange or green sticks, respectively. Ligands are shown as ball-and-stick models. Note the acidic platform residue Asp198, the hydrophilic residues below the ligands, and the hydrophobic residues forming the lid above. The schematic was produced using PyMOL (PyMOL Molecular Graphics System, Version 3.0 Schrödinger, LLC). (E and F) Schematic representation of CqsR_p-L-alaninol and CqsR_p-serinol binding. CqsR_p and L-alaninol are depicted as orange and gray bonds, respectively, in panel E. CqsR_p and serinol are depicted as green and gray bonds, respectively, in panel F. Hydrogen bonds are depicted as black dashed lines. Hydrophobic contacts are depicted as lines radiating from the semicircles and spheres. The schematic was produced using LigPlot+ [35].

https://doi.org/10.1371/journal.ppat.1013447.g005

The X-ray crystal structures of CqsR_p-ethanolamine, CqsR_p-L-alaninol, and CqsR_p-serinol in combination with the thermodynamic binding studies of CqsR_p to ethanolamine and ethanolamine analogs points to specific requirements for ligand binding to CqsR. Namely, the presence of both an amine positioned proximal for coordination by Asp171/Tyr169, and a hydroxyl oriented towards Thr196 are required for tight binding. Both of these key heteroatom functional groups of the ligand are chelated by Asp198, supporting the observed preference for a ligand with a two-carbon linker between these atoms. Furthermore, the structures and binding studies show where structural modifications in the ligand are tolerated. This is mainly in the observed binding of small α-amino substituents, including methyl and hydroxymethyl, as represented by the ligands L-alaninol and serinol, respectively. Weak CqsR binding is observed for ethylenediamine, in which the hydroxyl of ethanolamine bound by Thr196 is replaced by an amine, tris in which an additional α-amino substituent is incorporated, and both 2-amino-1-butanol and (S)-(+)-2-amino-1-butanol, in which an ethyl functional group replaces the L-alaninol methyl substituent (Fig 4). No binding is observed for ligands in which the amine functionality of the ligand is modified as in glycerol, ethylene glycol, and propanol (Fig 4). Furthermore, replacement of the hydroxyl group with a carboxylic acid (i.e., glycine and alanine), thiol (cysteamine), amide (glycinamide) or removal of this functionality (i.e., propylamine and pentan-3-amine) is not tolerated (Fig 4).

Structural conservation of the ethanolamine binding site

In addition to CqsR, three other periplasmic ethanolamine-binding receptors have been identified. Two of these proteins, Desulfovibrio inopinatus WP_027185430.1 and Peptoclostridium littorale WP_052635864.1, previously referred to as R3 and R7, respectively, were shown to bind ethanolamine in an isothermal titration assay [37]. In line with the binding affinity of CqsR_p and ethanolamine, R3 bound ethanolamine with K_d= 2.3 ± 0.1 μM and R7 bound ethanolamine with K_d= 1.7 ± 0.02 μM. The R3 and R7 periplasmic domains (R3_p and R7_p) are Cache domains whose structures are predicted to be similar to CqsR_p as determined by sequence analysis and structure prediction using AlphaFold 3 (S6 Fig) [38]. Based on sequence conservation in the ethanolamine binding site, we predict that ethanolamine binding to R3 and R7 will be structurally similar to that of CqsR (S6 Fig). Namely the use of a platform acidic residue (R3-Asp186 and R7-Asp207) to bind the ethanolamine amine and hydroxyl, and the use of aromatic side chains to form a hydrophobic lid above the ethanolamine alkane atoms (S6 Fig).

In addition, the solute-binding protein Csal_0678 from Chromohalobacter salexigens binds ethanolamine, and the Csal_0678 X-ray crystal structure was determined in complex with ethanolamine to 1.40 Å resolution (PDB ID 4UAB) [39]. While there is no recognizable similarity between the tertiary structures of Csal_0678 and CqsR, R3, and R7, i.e., Csal_0678 is not a Cache domain, their ethanolamine binding sites are remarkably similar (S6 Fig). More specifically, Csal_0678 uses a platform acidic residue (Glu220) to coordinate the ethanolamine amine and hydroxyl in an analogous manner to the CqsR, R3, and R7 platform Asp and surrounds the ethanolamine alkane atoms with aromatic side chains.

Targeted mutagenesis of the CqsR ligand-binding site

Guided by the ligand-bound CqsR_p structures, we predicted that Trp151, Tyr169, Asp171, Thr196, and Asp198 are important for ethanolamine binding to CqsR. To test the role of these residues in CqsR signal recognition inside the bacterial cell, we introduced Trp151Ala (W151A), Tyr169Ala (Y169A), Asp171Ala (D171A), Thr196Ala (T196A), or Asp198Asn (D198N) mutations into CqsR of a strain where the other three QS receptors are absent. This strain also carries a P_qrr4-luxCDABE transcriptional reporter so we can measure bioluminescence at different cell densities to monitor quorum-sensing (qrr4 transcription) in response to different concentrations of exogenously added ethanolamine.

Without any exogenously added ethanolamine, the strain with WT CqsR showed a high level of qrr4 transcription at low cell density (OD₆₀₀ ~ 0.2-0.4) and a low level of qrr4 transcription at high cell density (OD₆₀₀ > 0.8), as indicated by the increase and the subsequent decrease of bioluminescence during growth. The decrease of qrr4 transcription is due to the inhibition of CqsR activity by an unknown native autoinducer, i.e., an autoinducer produced by V. cholerae. In contrast, the level of qrr4 transcription was reduced in a dose-dependent manner by exogenously added ethanolamine (Figs 6, S7 and S8).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Response to exogenously added ethanolamine by different CqsR variants in V. cholerae.

Whole-cell bioluminescence assays were performed using a ∆cqsS ∆luxPQ ∆cqsR ∆vpsS strain with different CqsR variants expressed from a plasmid. These strains also contain a P_qrr4-luxCDABE reporter to measure the transcription of qrr4. Ethanolamine was added at the final concentrations as shown. Curves of the same color represent bacterial cultures grown with the same concentration of ethanolamine. For comparison, the two dotted lines on each graph show the maximum light production level of WT in the absence of ethanolamine and the minimal light production level of WT in the presence of the highest amount of ethanolamine at low cell density (OD₆₀₀ ~ 0.2). Representative results are shown with technical duplicates. The experiments have been repeated at least three times. In S7 Fig, these data for mutant CqsR are presented in comparison to the strain containing WT CqsR using a Y-axis scale that improves the visibility of the dose-response curves. In S8 Fig, biological replicates with technical triplicates are shown.

https://doi.org/10.1371/journal.ppat.1013447.g006

The qrr4 transcription pattern in the CqsR-Y169A or -T196A mutant was slightly different from that in the WT. In the absence of exogenously added ethanolamine, these two strains in general produced higher bioluminescence, suggesting the intrinsic CqsR kinase activity in these mutants is higher, possibly due to the inability of the natively made autoinducer to bind to the mutant receptors. However, in both mutants, qrr4 transcription was reduced to a level comparable to that of the WT by exogenously added ethanolamine (Figs 6, S7 and S8). Similarly, CqsR-W151A, -D171A, and -D198N produced higher bioluminescence than WT in the absence of exogenously added ethanolamine. However, CqsR-W151A and -D171A displayed an intermediate phenotype, responding to the highest concentrations of ethanolamine but never reducing bioluminescence comparable to that of the WT CqsR strain. In contrast, CqsR-D198N was completely irresponsive to ethanolamine.

These results are consistent with those of our previous genetic screen, where CqsR-D171V and CqsR-D198V resulted in a dramatic loss of signaling sensitivity to the native CqsR autoinducer [30]. The results of our site-directed and random mutagenesis confirm the in vivo relevance of the ligand-bound CqsR_p X-ray crystal structures and suggest that the ethanolamine-binding residues are also important for the recognition of the native autoinducer.

Insights into CqsR regulation by ethanolamine, serinol, and L-alaninol

Histidine kinases are dimers whose interaction is mediated by their cytoplasmic dimerization and histidine kinase domains (DHp domains) [40–43]. Analytical size-exclusion chromatography of CqsR_p-ethanolamine and apo CqsR_p-D198N showed that they are monomeric in solution (Fig 7). Consistent with these results, there is only one protomer in the CqsR_p-ethanolamine, CqsR_p-serinol, and CqsR_p-L-alaninol crystallographic asymmetric units. The application of crystallographic symmetry revealed that none of the symmetry-mates formed physiologically relevant dimers, i.e., interactions positioning their N- and C-termini on the same face in a manner consistent with CqsR_p dimer formation in the V. cholerae inner membrane. In addition, while the CqsR_p-D198N crystallographic asymmetric unit was dimeric, the protomers were anti-parallel, which is inconsistent with a biological arrangement. Together, the solution studies and crystallographic analysis suggest that signaling through CqsR is not regulated by ligand-induced shifts in its monomer-dimer equilibrium. Thus, we propose CqsR activity is regulated through the individual monomers within a presumed signaling dimer. It is important to note that this in vitro analysis was carried out in the absence of the CqsR transmembrane domain and membrane. It is possible that in the presence of these elements (or the V. cholerae autoinducer), CqsR_p forms interdimer contacts.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. CqsR_p-ethanolamine and CqsR_p-D198N are monomeric in solution.

Size exclusion chromatography of CqsR_p-ethanolamine and CqsR_p-D198N, with peak volumes indicated. CqsR_p-ethanolamine (MW_theor = 24.3 kD, MW_exp = 22.8 kD). CqsR_p-D198N (MW_theor = 24.2 kD, MW_exp = 19.3 kD). Vertical lines above the absorbance traces indicate protein size standards elution volumes (peak positions). Absorbance was measured at 280 nm and reported as milliabsorbance units (mAU).

https://doi.org/10.1371/journal.ppat.1013447.g007

To identify ethanolamine-induced conformational changes in CqsR_p monomers, we aligned the α1 helices (residues 44–77) of CqsR_p-ethanolamine and CqsR_p-D198N (Fig 8 and S1 Movie). Within Cache monomers, this α1 helix generally undergoes little if any conformational change upon ligand-binding, and it serves as a post against which both the membrane distal and proximal Cache domains move. When ethanolamine binds to the membrane-distal Cache domain, it triggers a reorientation of residues in the binding site (Fig 8 and S1 Movie). Coincident with this movement surrounding the ligand that compacts the membrane distal Cache domain, the membrane-proximal domain rotates anticlockwise relative helix α1. The movement of the membrane-proximal domain would cause a shift in the connected transmembrane region. The absence of the transmembrane domain from the crystal structures precludes our ability to describe this change, but the relative movement of the transmembrane domains, either downward along the z-axis (like a piston) and/or a combination of xy shifts (like scissors), has been proposed to regulate the cytoplasmic histidine kinase activity of histidine kinase receptor dimers [44].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 8. Ethanolamine-induced CqsR_p conformational changes.

Two different views of apo CqsR_p-D198N (gray cartoon) and CqsR_p-ethanolamine (magenta cartoon) aligned using their α1 helices (residues 44-77). Ethanolamine binding to the membrane distal Cache domain triggers its compaction, and the residues corresponding to helices α3 and α4 become ordered. These conformational changes in the membrane distal Cache domain result in the anticlockwise rotation of the membrane proximal Cache domain, which is more apparent in S1 Movie. Ethanolamine is represented as spheres.

https://doi.org/10.1371/journal.ppat.1013447.g008

CqsR_p production

CqsR_p encoding for residues 44–260 from V. cholerae O1 El Tor was cloned into pTB146, resulting in pHis₆-SUMO-CqsR_44-260. His₆-SUMO-CqsR_44-260 was overexpressed in E. coli BL21(DE3) by growing the cells in LB medium containing 100 μg/mL ampicillin at 37°C until reaching an OD₆₀₀ of 0.3. The temperature was then reduced to 18°C, and the culture was grown for 1 h prior to induction with 1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) for 18–20 hours at 18°C. The cells were collected by centrifugation at 4,000 rpm and resuspended in lysis Buffer A (300 mM NaCl, 50 mM HEPES (pH 8.0)) supplemented with 10% glycerol (vol/vol), 2 μg/mL DNase, and 2 μg/mL RNase. The resuspended cells were passed through a French Press at 14,000 psi and then centrifuged at 25,000 rpm for 30 min. The supernatant was incubated with His60 Ni Superflow resin (Takara) then washed with Buffer A. His₆-SUMO-CqsR_44-260 was eluted with Buffer A containing increasing concentrations of imidazole (50 mM – 500 mM). Protein purity was assessed using SDS-PAGE. Fractions containing His₆-SUMO-CqsR_44-260 were mixed with Ulp1 and dialyzed against Buffer A supplemented with 0.01% Triton. The dialysate was loaded onto a His60 Ni Superflow column to separate cleaved CqsR_44-260 from His₆-SUMO. Fractions containing cleaved CqsR_44-260 were pooled and concentrated using a 10 kDa MWCO centrifugal filter at 3,500 rpm. Concentrated CqsR_44-260 was passed through a 0.22 μM filter and loaded onto a Superdex200 (16/70) size exclusion column (Cytiva) equilibrated with Buffer B (150 mM NaCl, 20 mM HEPES (pH 8.0)). Fractions containing CqsR_44-260 were concentrated using a 10 kDa MWCO centrifugal filter and stored at -80°C.

Generation and purification of CqsR_p-D198N

Substitution mutations in pHis₆-SUMO-CqsR_44-260 were generated using the Q5 Site-Directed Mutagenesis Kit (New England Biolabs) and the primer pairs indicated in S4 Table. DNA sequences were confirmed by whole plasmid sequencing using Oxford Nanopore Technology with custom analysis and annotation (Plasmidsaurus). CqsR_p-D198N was purified using the protocol described for wild-type CqsR_p.

CqsR_p-ethanolamine, CqsR_p-L-alaninol, CqsR_p-serinol, and CqsR_p-D198N crystallization, X-ray diffraction data collection, and structure solution

Crystals of CqsR_p in complex with ethanolamine were obtained via the vapor diffusion method by first incubating 206.7 μM CqsR_p with 1.0 mM ethanolamine on ice for approximately 5 min. 1.0 μL of the CqsR_p-ethanolamine mixture was then mixed with 1.0 μL of mother liquor containing 1.8 M lithium chloride, 0.1 M bicine (pH 9.0), and 10% (v/v) PEG 6000 (S1 Table) at 20°C to generate a 2.0 μL hanging drop. Prior to X-ray diffraction data collection, crystals were moved to a solution of mother liquor supplemented with 12% PEG 400 and 10.0 mM ethanolamine. X-ray diffraction data for CqsR_p-ethanolamine were collected using single crystals mounted in nylon loops, and data collection occurred in a stream of dry N₂ at 100 K. X-ray diffraction data for CqsR_p-ethanolamine were collected at the Standford Synchrotron Radiation Lightsource (SSRL) beamline 9–2 at 0.97946 Å with a Pixel Dectris Pilatus 6M detector.

Crystals of CqsR_p in complex with L-alaninol were obtained via the vapor diffusion method by first incubating 206.7 μM CqsR_p with 10.0 mM L-alaninol on ice for approximately 5 min. 1.0 μL of the CqsR_p-L-alaninol mixture was then mixed with 1.0 μL of mother liquor containing 1.8 M lithium chloride, 0.1 M bicine (pH 9.0), and 10% (v/v) PEG 6000 (S1 Table) at 20°C to generate a 2.0 μL hanging drop. Prior to X-ray diffraction data, crystals were moved to a solution of mother liquor supplemented with 18% PEG 400 and 10.0 mM L-alaninol. X-ray diffraction data for the CqsR_p-L-alaninol were collected using single crystals mounted in nylon loops and data collection occurred in a stream of dry N₂ at 100 K. X-ray diffraction data for CqsR_p-L-alaninol was collected in-house using the Rigaku Synergy S at 1.540562 Å with a Rigaku HyPix-6000HE detector.

Crystals of CqsR_p in complex with serinol were obtained via the vapor diffusion method with 1.0 μL of 206.7 μM CqsR_p combined 1:1 with 1.0 μL mother liquor containing 1.8 M lithium chloride, 0.1 M bicine (pH 9.0), and 10% (v/v) PEG 6000 (S1 Table) at 20°C to generate a 2.0 μL hanging drop. Prior to X-ray diffraction data, crystals were moved to a solution of mother liquor supplemented with 15% PEG 400 and 10.0 mM serinol. X-ray diffraction data for CqsR_p-serinol were collected using single crystals mounted in nylon loops, and data collection occurred in a stream of dry N₂ at 100 K. X-ray diffraction data for CqsR_p-serinol were collected in-house using a Rigaku Synergy S at 1.540562 Å with a Rigaku HyPix-6000HE detector.

Crystals of CqsR_p-D198N were obtained via the vapor diffusion method with 1.0 μL of 206.7 μM CqsR_p-D198N combined 1:1 with 1.0 μL mother liquor containing 2.0 M sodium formate and 2% (v/v) glycerol (S1 Table) at 20°C to generate a 2.0 μL hanging drop. Prior to collecting X-ray diffraction data, crystals were moved to a solution of mother liquor supplemented with 25% ethylene glycol. X-ray diffraction data for CqsR_p-D198N were collected using single crystals mounted in nylon loops, and data collection occurred in a stream of dry N₂ at 100 K. X-ray diffraction data for CqsR_p-D198N were collected in-house using a Rigaku Synergy S at 1.540562 Å with a Rigaku HyPix-6000HE detector.

X-ray diffraction data for CqsR_p-ethanolamine were processed using HKL 3000 [65]. X-ray diffraction data for CqsR_p-L-alaninol CqsR_p-serinol, and CqsR_p-D198N were processed using CrysAlis PRO (Agilent (2014). Agilent Technologies Ltd, Yarnton, Oxfordshire, England).

Initial crystallographic phases were determined by molecular replacement using Phaser [66] and a CqsR_p search model generated using AlphaFold 3 [38] or CqsR_p from the CqsR_p-ethanolamine structure. Manual model building was carried out in Coot [34] and refinement was performed in phenix.refine [67]. Initial model refinement included either automated model building or simulated annealing followed by rigid body, individual atomic coordinate, and individual B-factor refinement. Later rounds of refinement employed individual atomic coordinate, individual B-factor, and TLS refinement. TLS groups were identified using the automatic TLS selection algorithm in the Phenix selection editor. During the final rounds of refinement, the stereochemistry and ADP weights were optimized. Insufficient electron density was observed for the following residues in the flexible regions of the protein structures, and they were omitted from the model: CqsR_p-ethanolamine 128–130; CqsR_p-L-alaninol 128–130; CqsR_p-serinol 111–112, 128; CqsR_p-D198N chain A 44–45, 110–117, 128–131, 140–144, 235–237, 260, and chain B 110–115, 128–131, 141–144, 235–237, 260. Ramachandran statistics were calculated in Molprobity [68]. Molecular graphics were produced with PyMOL (PyMOL Molecular Graphics System, Version 3.0 Schrödinger, LLC).

Microscale thermophoresis

MST ligand-binding experiments were performed using wild-type CqsR_p purified as described above. CqsR_p was labeled using the Protein Labeling Kit RED-NHS 2^nd Generation (NanoTemper Technologies). Absorbance readings at 205 nm, 280 nm, and 650 nm were used to determine the degree of labeling and protein concentration. 100.0 nM total CqsR_p (18.8 nM ligand-free CqsR_p) in Buffer B supplemented with 0.2% Pluronic acid was mixed with ligand and incubated in the dark at room temperature for 30–45 min. Samples were then transferred to standard treated capillaries (NanoTemper Technologies) and analyzed using a NanoTemper Monolith NT.115 Blue/Red instrument at 60% LED power and medium MST power.

CqsR_p LC-MS/MS

A sample of 103.3 μM purified CqsR_p and a 103.3 μM sample of CqsR_p-D198N in buffer (150 mM NaCl, 20 mM HEPES, pH 8.0) were denatured at 70°C for 3 minutes and then centrifuged at 13,000 rpm for 10 minutes at 23ºC. Ethanolamine present in the supernatant from each sample was derivatized and analyzed as described previously with modifications [30]. Briefly, 50.0 μL of the resulting supernatant was mixed with 25.0 μL of 0.1 M NaHCO₃/NaCO₂ (pH = 9.5), 25.0 μL of acetonitrile, and 500.0 μL of freshly prepared dansyl chloride [1.8 mg/mL in acetonitrile]. The sample was heated at 60°C for 60 minutes. Then, 20.0 μL of 0.250 M NaOH was introduced and heated for an additional 10 minutes at 60°C. Afterwards, 100.0 μL of 0.425 M formic acid was incorporated and diluted 1:1 with 720.0 μL of 0.1% formic acid in water. The samples were finally filtered through a 0.22 μm PTFE syringe filter.

LC-MS/MS analysis of dansyl monoethanolamine was performed using a 6430 Triple Quadrupole LC/MS (Agilent Technologies, Santa Clara, CA, USA). A 2.0 μL aliquot of the derivatized product was injected into a Hypersil BDS C8 HPLC Column (50 x 2.1 mm x 5 μm; Thermo Fisher Scientific, Waltham, MA, USA). The mobile phases used were: A, 0.1% formic acid in HPLC grade water and B, 0.1% formic acid in acetonitrile. The column was maintained at ambient temperature with a flow rate of 0.200 mL/min. A gradient was run from 5% B to 30% B over 8 min, increased to 90% B over 1 min, and held for 2 min. The gradient returned to 5% B over 1 min and was equilibrated at 5% B for 2 min (14 min run time). Acquisition was performed in positive ion mode and the collision energy was set to 20 eV.

A MS/MS experiment was conducted to determine the fragmentation pattern and to select the appropriate product ions for monitoring. Multiple reaction monitoring (MRM) was performed using the [M + H]+ precursor ion of dansyl monoethanolamine at m/z 295 with unit mass resolution in the first quadrupole. MRM transitions of the precursor/product ions were 295/280, 295/171, and 295/157 (m/z), corresponding to the fragmentation of the precursor ion. These transitions were used to detect product ions derived from dansyl monoethanolamine. The target eluted at approximately 4.12 min for CqsR_p and 4.71 min for CqsR_p-D198N. Ethanolamine concentration in each sample was inferred from ion counts using a series of pure ethanolamine standards prepared at the time of each analysis.

Strains, media and culture conditions

All V. cholerae strains used in this study were derived from C6706str2, a streptomycin-resistant isolate of C6706 (O1 El Tor) [69]. The lac::P_qrr4-luxCDABE transcriptional fusion was integrated into the genome of the strain WN5870 lacking the four quorum sensing receptors (∆cqsS ∆luxPQ ∆cqsR ∆vpsS) [30] using the plasmid pKAS32 as described [70]. To this reporter strain, CqsR or its variant was introduced on the plasmid pEVS143 under control of a P_tac promoter as previous described [30]. These CqsR variants, each with a single change at a particular amino acid residue, were constructed using site-directed mutagenesis with the plasmid pEVS143-CqsR^WT as the template. A complete list of strains and primers used in this study are provided in S5 and S6 Tables.

V. cholerae and E. coli cultures were grown with aeration in Luria-Bertani (LB) broth at 30°C and 37°C, respectively. Unless specified, media was supplemented with streptomycin (Sm, 100.0 μg/ml), tetracycline (Tet, 5 μg/mL), ampicillin (Amp, 100 μg/mL), kanamycin (Kan, 100 μg/mL), chloramphenicol (Cm, 5 μg/mL) and polymyxin B (Pb, 50 U/mL) when appropriate. Stock solution of ethanolamine (Sigma-Aldrich) was prepared in water and used as shown in this study.

Bioluminescence assays

Bioluminescence assays to measure qrr4 transcription in different V. cholerae reporter strains were performed as described previously with slight modifications [11,30]. Single colonies of each strain were grown in LB broth containing appropriate antibiotics for overnight at 30°C with shaking. The overnight culture was further diluted 1000-fold in LB with antibiotics and 10.0 μM IPTG, and 200.0 μL of the diluted culture was added to each well of a black 96-well microplate with a clear bottom. Ethanolamine stocks (100X) of various concentrations was prepared in water and 2.0 μL was added to each well containing the diluted bacterial culture. Mineral oil (50.0 μL) was added to each well to prevent evaporation without a lid cover. OD₆₀₀ and light production were measured every 30 mins for 20 hrs with shaking using a BioTek Synergy HT Plate Reader.