Crystal structure of an inactive variant of the quorum-sensing master regulator HapR from the protease-deficient non-O1, non-O139 Vibrio cholerae strain V2

The crystal structure of an inactive variant of the the quorum-sensing master transcription regulator HapR from Vibrio cholerae suggests that its inactivity is owing to steric clashes and charge repulsion with the phosphate backbone of DNA.


Introduction
The acute diarrheal disease cholera is caused by ingesting food or water contaminated with the Gram-negative bacterium Vibrio cholerae. The world is currently experiencing a seventh global cholera pandemic, which began in 1961. In 2015, 42 countries reported 172 454 cases and 1304 deaths. However, it is estimated that the actual number of cholera cases is between 1.3 and 4 million per year and that 21 000-143 000 die from the disease each year worldwide (World Health Organization, 2016).
Like many bacteria, V. cholerae uses quorum sensing to regulate gene expression in response to an increase in cell density (Miller & Bassler, 2001). V. cholerae secretes the autoinducers CAI-1 and AI-2 . As the cell density increases, the extracellular concentration of these autoinducers also increases. At low cell density, the response regulator LuxO is phosphorylated by the sensor kinases CqsS and LuxP (Higgins et al., 2007). Phosphorylated LuxO activates the transcription of four small regulatory RNAs that inhibit the translation of the quorum-sensing master regulator HapR (Lenz et al., 2004). At high cell density, CqsA and LuxS detect CAI-1 and AI-2, respectively, which leads to the dephosphorylation of LuxO and the production of HapR. HapR regulates the expression of a large number of genes in V. cholerae: it activates the expression of hemagglutinin protease (Silva & Benitez, 2004), enhances the stress response (Joelsson et al., 2007), enhances predation-driven persistence (Matz et al., 2005), promotes chitin-induced competence (Meibom et al., 2005), regulates hcp expression (Ishikawa et al., 2009), negatively regulates virulence-gene expression by repressing the expression of AphA (Zhu et al., 2002;Kovacikova & Skorupski, 2002;Lin et al., 2007) and represses biofilm formation by repressing vpsR (Hammer & Bassler, 2003).
HapR is a member of the TetR family of transcriptional regulators (Jobling & Holmes, 1997). The crystal structure of wild-type HapR revealed that the protein is entirely -helical and contains an N-terminal helix-turn-helix DNA-binding domain and a C-terminal dimerization/regulatory domain typical of TetR-family members (De Silva et al., 2007). Within each regulatory domain is an amphipathic cavity that may serve as a binding pocket for a yet to be identified ligand.
A surprising number of epidemic-causing O1/O139 strains as well as non-O1/non-O139 strains of V. cholerae isolated globally have been found to have dysfunctional quorumsensing systems (Joelsson et al., 2006;Talyzina et al., 2009;Wang et al., 2011). Of these, several have mutations in hapR. The classical strain O395 and the El Tor strain N16961 both have frameshift mutations that place a premature stop codon upstream of the C-terminal dimerization domain of HapR. A portion of the dimerization domain of HapR is deleted in strain SG1. Strains MO10, 857 and MAK757 have one, two and seven point mutations in hapR, respectively. Strain MDO14-T completely lacks hapR.
The protease-deficient V. cholerae serotype O37 strain V2 was isolated in Calcutta, India in 1989. Strain V2 was recently found to contain a glycine-to-aspartate substitution at position 39 within the hinge region between the DNA-binding and dimerization domains of HapR (HapR V2 ; Dongre et al., 2011). In their study, Dongre and coworkers showed by EMSA that HapR V2 was defective in DNA-binding activity. Size-exclusion chromatography and circular dichroism revealed no significant structural differences between normal and variant HapR. Furthermore, Guinier analysis and indirect Fourier transformation of small-angle X-ray scattering (SAXS) indicated only a slight difference in shape. However, structural reconstruction using the SAXS data suggested that the arrangement of the DNA-binding domains of the variant HapR was altered. To gain further insight into the functional role of the hinge region of HapR and the structural consequences of a substitution of aspartate for glycine at position 39, we determined the crystal structure of HapR V2 to a resolution of 2.1 Å . The structure suggests that the aspartate located in the hinge region of HapR V2 would sterically clash with and electro-statically repel DNA, preventing the binding and the regulation of genes controlled by the quorum-sensing system of V. cholerae.

Expression and purification
The ORF for HapR V2 was cloned into pET-15b to generate thrombin-cleavable N-terminally six-His-tagged HapR V2 , as described previously by Dongre et al. (2011). HapR V2 was expressed in Escherichia coli BL21(DE3) cells induced by autoinduction in ZYM-5052 medium overnight at 20 C (Studier, 2005). The cells were lysed in 20 mM Tris-HCl pH 8, 100 mM NaCl by sonication at 4 C and centrifuged at 120 000g for 30 min. The supernatant was filtered using a 0.45 mm filter and loaded onto a GE HisTrap FF column using an Ä KTAexplorer FPLC system. The column was eluted with a linear gradient of 40-500 mM imidazole and a single peak was collected. The protein was further purified using a GE SP FF cation-exchange column and a Superdex S75 16/600 sizeexclusion column.

Crystallization
Purified HapR V2 was concentrated to 5 mg ml À1 using Amicon Ultra centrifugal filter units. Crystallization conditions were screened by the sitting-drop vapor-diffusion method. Diffraction-quality single crystals were obtained by mixing

Data collection and processing
X-ray diffraction data were collected on beamline X6A at the National Synchrotron Light Source (NSLS), Brookhaven National Laboratory, Upton, New York, USA. A 2.1 Å resolution data set of 360 frames with an oscillation range of 0.5 was collected at a wavelength of 1.000 Å with 15 s exposures at 100 K. The crystal-to-detector distance was 220 mm. The data set was indexed, integrated, scaled and merged using XDS (Kabsch, 1993). Data-collection statistics are shown in Table 1.

Structure solution and refinement
The reflection file was converted and R free flags were set (7.89% of unique reflections) using the PHENIX reflectionfile editor (Adams et al., 2002). The Matthews coefficient was calculated and it was determined that the asymmetric unit contained a single dimer of HapR V2 . The structure of HapR V2 was solved by molecular replacement with PHENIX Phaser-MR (McCoy et al., 2007) using wild-type HapR (PDB entry 2pbx; De Silva et al., 2007) as the search model. Multiple rounds of refinement were carried out using Coot and phenix.refine (Emsley & Cowtan, 2004;Afonine et al., 2012). Refinement statistics are shown in Table 1.

Structural alignments and modeling
All structural alignments, modeling and distance measurements were performed with the PyMOL molecular-graphics system (DeLano, 2002).

Structure of HapR V2
The crystal structure of HapR V2 was refined to a resolution of 2.1 Å (Fig. 1). There is one homodimer of HapR V2 in the asymmetric unit. The two subunits in each dimer are related by twofold noncrystallographic symmetry. As in wild-type    HapR, the structure of HapR V2 is entirely -helical. The first three helices of each monomer form a helix-turn-helix DNAbinding domain. Helices 4-9 form the dimerization/regulatory domain.

Alignment of HapR V2 with wild-type HapR
Although previous structural reconstructions using smallangle X-ray scattering data have suggested that the DNAbinding domain of HapR V2 adopts an altered conformation relative to the wild type, the crystal structure of HapR V2 reveals that there are no significant structural differences between HapR V2 and wild-type HapR (Fig. 2a). Alignment of the HapR V2 dimer with wild-type HapR results in an r.m.s.d. of 0.448 Å for 381 C atoms. Phe55, which has been shown to be necessary for DNA binding (De Silva et al., 2007), is in an identical position to that in wild-type HapR (Fig. 2b).
Furthermore, residues within the putative ligand-binding pocket of HapR V2 are in the same positions as in the wild type (Fig. 2c).

Alignment of HapR V2 with the QacR-DNA complex
In order to gain further insight into the reason that HapR V2 is unable to bind DNA, the HapR V2 structure was aligned with that of the Staphylococcus aureus multidrug-binding transcriptional repressor QacR in complex with its operator DNA     ( Fig. 4a). HapR V2 aligns with QacR with an r.m.s.d. of 2.3 Å for 284 C atoms. The alignment positions the DNA-binding helices within adjacent major grooves of the DNA double helix. The C atoms of Phe55 in each subunit of the HapR V2 dimer are 39.7 Å apart, which is only 2.8 Å further apart than the C atoms of Tyr40 at the analogous positions in QacR. Interestingly, the positioning of HapR V2 on DNA revealed that the carboxyl side chain of Asp39 may both sterically clash with and electrostatically repel the phosphate backbone of DNA, possibly explaining the inability of this variant to bind DNA (Fig. 4b).
3.5. Electrostatic surface potential of HapR V2 versus the wild type A comparison of the electrostatic surface potential of HapR V2 with that of wild-type HapR revealed only subtle differences in the positions of charged residues (Fig. 5). However, the electrostatic surface of HapR V2 positioned on DNA by alignment with QacR shows that the negatively charged surface of Asp39 would overlap with the DNA backbone if bound (Fig. 5b).

Conclusion
Given the number of V. cholerae isolates that have been found with nonfunctional quorum-sensing systems, it can be assumed that the loss confers some advantage to the bacterium. The classical pandemic serotype O1 strain O395 and the El Tor strain N16961 both have mutations in the quorum-sensing master regulator HapR. The nonfunctional HapR from V. cholerae serotype O37 strain V2, isolated in Calcutta, India, bears a substitution of aspartate for glycine at position 39, which is within the hinge region between the DNA-binding and regulatory domains. The crystal structure of HapR V2 suggests that the carboxylate side chain of Asp39 would clash with and electrostatically repel the phosphate backbone of DNA, preventing DNA binding by HapR V2 and therefore the regulation of quorum-controlled genes. Electrostatic surfaces of wild-type HapR (a) and HapR V2 (b) positioned on DNA as aligned with the QacR-DNA structure. Positively charged surface is colored blue; negatively charged surface is colored red. Yellow arrows indicate where the side chain of Asp39 in HapR V2 would overlap with the phosphate backbone of DNA.