Structural insights into DNA binding domain of vancomycin‐resistance‐associated response regulator in complex with its promoter DNA from Staphylococcus aureus

Abstract

In Staphylococcus aureus, vancomycin‐resistance‐associated response regulator (VraR) is a part of the VraSR two‐component system, which is responsible for activating a cell wall‐stress stimulon in response to an antibiotic that inhibits cell wall formation. Two VraR‐binding sites have been identified: R1 and R2 in the vraSR operon control region. However, the binding of VraR to a promoter DNA enhancing downstream gene expression remains unclear. VraR contains a conserved N‐terminal receiver domain (VraR_N) connected to a C‐terminal DNA binding domain (VraR_C) with a flexible linker. Here, we present the crystal structure of VraR_C alone and in complex with R1‐DNA in 1.87‐ and 2.0‐Å resolution, respectively. VraR_C consisting of four α‐helices forms a dimer when interacting with R1‐DNA. In the VraR_C–DNA complex structure, Mg²⁺ ion is bound to Asp194. Biolayer interferometry experiments revealed that the addition of Mg²⁺ to VraR_C enhanced its DNA binding affinity by eightfold. In addition, interpretation of NMR titrations between VraR_C with R1‐ and R2‐DNA revealed the essential residues that might play a crucial role in interacting with DNA of the vraSR operon. The structural information could help in designing and screening potential therapeutics/inhibitors to deal with antibiotic‐resistant S. aureus via targeting VraR.

Short abstract

PDB Code(s): 7VE4, 7VE5 and 7VE6

1. INTRODUCTION

The emergence of antibiotic‐resistance mechanisms from infection‐causing microbes is a threat to the treatment of diseases. ¹ , ² , ³ Among antibiotic‐resistant microorganisms, the group of so‐called ESKAPE bugs—Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter species—was found responsible for severe nosocomial infections. ⁴ , ⁵ , ⁶ In particular, methicillin‐resistant S. aureus (MRSA) infection is often implicated in most cases of hospitalized infections. ⁷ , ⁸ , ⁹ , ¹⁰ Specifically, the Mu50 MRSA strain is vancomycin‐resistant. ¹¹ , ¹² The introduction of antibiotics leads to a cell wall stress stimulon in vancomycin‐resistant S. aureus infection. ¹³ , ¹⁴

To cope with environmental stress, these bacteria sense the stress and respond, thereby becoming more resilient. One major contributor to signaling mechanisms acts via a two‐component signal (TCS) transduction system. ¹⁵ , ¹⁶ The TCS is based on a conserved phosphotransfer pathway between histidine kinase (HK) and its response regulator (RR). Typically, the response regulator contains a receiver domain (RD) and a DNA‐binding domain (DBD) connected with a linker. ¹⁷ , ¹⁸ The autophosphorylation of HK leads to an active phosphoryl group, which is transferred to the conserved Asp residue on the RD of RR. ¹⁷ Furthermore, phosphorylation of RRs often leads to structural changes and enhances binding to its promoter DNA. ¹⁹ , ²⁰ , ²¹

In S. aureus, the vancomycin resistance‐associated sensor/regulator (VraSR) TCS comprises HK (VraS) and RR (VraR), which recognize and transduce cell wall stress. ¹⁴ , ²² , ²³ In addition to vancomycin resistance, the VraSR system was demonstrated as a “central regulation factor” for its involvement in regulating virulence. ²⁴ S. aureus VraR belongs to the NarL/FixJ family, which is characterized by a typical helix‐turn‐helix (HTH) structural motif at the C‐terminal DNA binding domain. ²⁵ The phosphorylated VraR binds to promoter DNA and increases the expression of the vraSR operon and the regulation of additional genes responsible for cell wall synthesis. ¹⁴ , ²⁰ , ²⁶ Two VraR‐binding sites, R1 and R2, in the vraSR operon control region have been identified. ²⁰ However, the detailed mechanism of VraR binding to promoter DNA for regulating downstream gene expression is still unclear.

The structural changes upon the activation of full‐length RRs have been moderately understood. ¹⁷ , ²⁷ Previous studies demonstrated that VraR predominantly exists as a monomer and is induced to form a dimer by phosphorylation, which allows for extended DNA recognition. ²⁰ , ²¹ , ²⁸ The crystal structure of VraR revealed a closed conformation between its RD and DBD in the un‐phosphorylated form, whereas the presentation of beryllofluoride (BeF₃ ⁻) for VraR activation leads to an open conformation and promotes dimerization. ²⁸ VraR phosphorylation leads to the opening of a hydrophobic pocket at the RD in which the Met13 residue of one protomer protrudes into its opposite protomer. ²⁸ A recent study showed that the substitution of a single residue Met13 with alanine greatly inhibited the dimer formation and suggested that any effort to hamper the phosphorylation‐induced dimerization can disrupt cell wall stress in S. aureus. ²⁹

Here, we report the crystal structure of VraR_C in complex with R1‐DNA elaborating clear and potent binding interactions. The complex structure shows a compact dimer between DBDs when binding to R1‐DNA. The positively charged surface (α9 helix) of VraR_C compliments the negatively charged DNA phosphodiester backbone. Also, the complex structure of VraR_C–DNA contains six Mg²⁺ ions, two of which coordinated with Asp194 of each subunit. Further analysis with biolayer interferometry assay (BLI) revealed that the presence of Mg²⁺ ion increased the binding affinity of VraR_C toward DNA. Our study revealed the atomic molecular interactions of VraR_C to promoter DNA with solid biophysical and biochemical evidence.

2. RESULTS

2.1. Characterization of VraR in solution

We constructed and purified the recombinant full‐length VraR (residues 1–209) and VraR_C (residues 138–209). From the elution profiles of size‐exclusion chromatography along with the calibrated standard protein markers, both VraR and VraR_C eluted as monomers with a molecular weight 23.56 and 8.33 kDa, respectively. The high purity of samples was assessed (> 95%) on Coomassie‐blue‐stained SDS‐polyacrylamide gel (Figure S1). A previous report showed that the phosphorylation of VraR was mimicked with BeF₃ ⁻ and led to a stable dimer. ²⁸ Here, we followed a similar procedure by incubating VraR with 5 mM BeF₃ ⁻ for activation and validated the dimer formation via analytical ultracentrifugation–sedimentation velocity (AUC‐SV) (Figure S2). The circular dichroism (CD) spectra for VraR_C, inactive and active forms of VraR in 20 mM sodium phosphate, pH 6.0, at 25°C showed well‐folded conformations with a typical far‐UV spectrum and negative signals at 222 and 208 nm and a positive signal at 193 nm, which indicates a classical helical structure (Figure S3). The thermal unfolding experiments were performed with CD spectroscopy for VraR_C, inactive and active VraR in 20 mM sodium phosphate, at pH 6.0, 7.0, and 8.0. VraR_C monitored at 222 nm was most stable at pH 6.0, with melting temperature 58.3°C (Table S1). A moderately higher thermal stability for active VraR at 52.17°C than the inactive VraR at 48.82°C agrees with previous findings. ³⁰

To understand the stoichiometry of both inactive and active VraR toward R1‐DNA DNA, we analyzed the complex formation by using AUC‐SV. Equimolar ratios of active VraR with R1‐DNA mixture displayed two peaks in a sedimentation coefficient [c(S)] distribution plot with 2.69 and 5.07 Svedberg (S) (Figure 1). The unbound R1‐DNA was observed in the first peak at 2.69 S, whereas the second peak at 5.07 S corresponds to ~64.5 kDa, which is close to the theoretical molecular weight of dimerized active VraR (47.12 kDa) in complex with R1‐DNA (18 kDa). Our finding for inactive VraR to R1‐DNA in AUC‐SV resulted in a peak at 2.7 S corresponding to ~23.3 kDa, indicating both inactive VraR and unbound DNA. However, the anticipated inactive VraR‐DNA complex at 5.2 S displayed a minor peak, suggesting a partial complex formation.

FIGURE 1.

AUC‐SV profiles showing sedimentation coefficient distribution. All the data were analyzed by using Sedfit for assessing molecular weights. (a) The c(S) distribution of unbound R1‐DNA is at 2.69 S and active VraR in complex with R1‐DNA had a peak shift at 5.07 S. (b) The c(S) distribution of inactive VraR and unbound R1‐DNA are at 2.7 S and a minor peak at 5.2 S for inactive VraR–DNA complex

2.2. Crystal structure of VraR_C and its dimerization with R1‐DNA complex

In this study, VraR_C was crystalized alone and in complex with its promoter DNA. The VraR_C crystal structure was solved to a resolution of 1.87 Å by the molecular replacement method with VraR as a model (PDB ID: 4GVP). VraR_C, consisting of four α‐helices designated α7, α8, α9 and α10 (Figure S4), was validated to have an identical conformation when aligned to the DBD of full‐length inactive VraR with root mean square deviation (RMSD) of 0.31 Å. To recognize the atomic basis for specific DNA recognition by VraR_C, we performed crystallization trials with various lengths of double‐stranded R1‐ and R2‐DNAs, including the blunt end or sticky end of one base (Table S2). Only R1‐DNA with 22 bp was successful in multiple crystallization trials with VraR_C. The co‐crystal structure of VraR_C in complex with R1‐DNA was solved to a resolution of 2.0 Å in space group‐R3 by using the molecular replacement method. All the crystallographic and refinement statistics are in Table S3. An asymmetric unit consists of two VraR_C subunits with electron density available from Glu143‐Ile208 and R1‐DNA nucleotides (5′‐AGACTAAAGTATGAACATC‐3′) near the protein‐bound region. In the VraR_C‐DNA complex, the dimerization interface of two α10 helices between VraR_C subunits allows for compact dimer formation when binding to R1‐DNA (Figure 2a). Specifically, three hydrogen bonds formed between residues Thr196‐Thr196, Gly162‐Gln197 and Gln197‐Gly162 among two α10 subunits of VraR_C.

FIGURE 2.

Structural overview of VraR_C in complex with promoter DNA. (a) Upstream and downstream binding of two VraR_C subunits binding to R1‐DNA are shown in blue and orange, respectively. Secondary structural elements on both the subunits are labeled and six Mg²⁺ ions are in green spheres. (b) Schematic diagram showing detailed interactions between two VraR_C molecules and DNA. Hydrogen bonds, electrostatic interactions are shown in green and brown dotted lines

Structural alignment between the VraR_C dimer from the R1‐DNA complex and domain‐swapped dimer of full‐length active VraR (PDB ID:4IF4) revealed high similarity, with an RMSD of 0.56 Å. However, the superimposition of a single subunit caused slight rotation in the second subunit (Figure 3). The distance of two Lys177 residues from the DNA‐bound VraR_C dimer was 31.6 Å as compared with 33.6 Å for the active VraR DBD, which suggests that both α9 helices move closer when interacting with DNA. Moreover, the buried surface area between the dimer interface of the VraR_C–DNA complex and active VraR DBD was 443.75 and 420.3 Å, respectively. Also, the superimposition of two VraR_C subunits from an asymmetric unit of the VraR_C–DNA complex had an RMSD of 0.09 Å. The above assessment showed no significant conformational changes within the VraR_C dimerization upon DNA binding.

FIGURE 3.

Superimposition of VraR_C from the R1‐DNA complex in yellow and activated VraR DBD (PDB ID: 4IF4) in light‐cyan. The Subunit‐B has slight rotation upon binding to DNA. Lys177 and Lys180 in the VraR_C dimer from α9 helices are involved in binding to DNA are shown as sticks. Distances between two Lys177 and two Lys180 residues within DNA‐bound VraR_C dimers were 2.2 and 1.0 Å closer than active VraR

2.3. Major and minor groove interactions of VraR_C –DNA complex

In the VraR_C–DNA complex structure, R1‐DNA presents a β‐form conformation with base stacking interactions between the two complementary DNA strands. Eleven nucleotides from the DNA duplex had hydrogen and salt bridge interactions with the VraR_C dimer (Figure 2b). Similar to the other complex structures of the NarL/FixJ family, both the α9 helices from each VraR_C subunit are inside the DNA major groove. Of six residues (Thr149, Arg151, Thr175, Thr178, Thr181, and His182) binding to the DNA major groove (T_5′T_6′C_7′A_8′), three threonine's (Thr175, Thr188, and Thr181) interact with C_7′ alone. Mutations of corresponding C_7′ and C_11′ to T reduced the binding affinities of inactive and active VraR by ~6‐ and threefold toward R1‐DNA. ²⁰

Four residues from each VraR_C subunit, Asn165, Lys180, Ser184, and Arg195, interact with the DNA minor groove (A₈G₉T₁₀ in one half‐site and A_15′G_16′T_17′ in the other half‐site). Ser184 binds to T₁₀ via a hydrogen bond with distance of 2.7 Å. Residue Asn165 of α8 contacts A₈ with a hydrogen bond (3.1 Å) along with two water‐mediated hydrogen bonds toward A₈ and G₉. Both oxygen atoms of G₉ phosphate bind to Arg195 with hydrogen‐bond and salt‐bridge interactions (2.7 and 2.8 Å). Furthermore, Lys180 bind to oxygen and nitrogen atoms of the G₉ base. A previous study showed a sevenfold decrease in binding affinity between inactive VraR and a sixfold decrease for active VraR toward R1‐DNA when nucleotides G₉ and C₄ were mutated to T²⁰. Since Lys180 and Arg195 form two bonds with G₉, these two residues might play a crucial role in recognizing R1‐DNA.

2.4. NMR study reveals fewer VraR_C residues interacting with R2‐DNA than R1‐DNA

A previous study presented the NMR backbone assignment of VraR_C from Leu144‐Gln209 with six missing residues (Glu150, Arg151, Asn165, Gln166, His173, and Ile176). ³⁰ By contrast, all the amide resonances except for Lys139 and Asn165 were unambiguously assigned for VraR_C in this study (Figure S5). Furthermore, triple‐labeled samples of full‐length VraR in both inactive and active forms were prepared, and 2D ¹H¹⁵N TROSY‐HSQC spectra were obtained. Increasing the protein concentration to >0.3 mM resulted in heavy precipitation, which hampered our trials for VraR backbone assignment. In comparison, the 2D ¹H¹⁵N TROSY‐HSQC spectra of active VraR showed more and well‐dispersed cross peaks than the inactive form (Figure 4). Accordingly, active VraR showed better solubility and stability than the inactive VraR at pH 6.0, which is consistent with the results of thermal denaturation experiments. The 2D ¹H¹⁵N TROSY‐HSQC spectra of VraR_C were overlaid on inactive and active VraR. Crucial residues from the C‐terminal domain (Glu191 and Gln204) responsible for closed interface with N‐terminal domain of inactive VraR and the residues attributed to DBD dimerization (Gly162, Thr196, and Gln197) in active VraR were declared on VraR_C assignment (Figure 4). Although the backbone assignment of the full‐length VraR is unobtainable, the overall cross peaks of VraR_C are similar to both active and inactive forms of VraR.

FIGURE 4.

(a) The overlaid 2D ¹H, ¹⁵N TROSY‐HSQC spectra were acquired from inactive VraR (blue) and VraR_C (red). Residues Gln204 and Glu191 of α10 helix that interact with the N‐terminal domain of inactive VraR are labeled on VraR_C. (b) The overlaid 2D ¹H, ¹⁵N TROSY‐HSQC spectra were acquired from active VraR (blue) and VraR_C (red). Residues Gly162, Thr196 and Gln197 involved in the DBD dimer of active VraR are labeled on VraR_C. All samples were prepared at pH 6.0 and acquired spectra at 298 K

We used NMR to analyze the binding interface of VraR_C toward both R1‐ and R2‐DNA by chemical shift perturbations (CSPs). The weighted ¹H–¹⁵N chemical shift differences were measured and charted from HSQC spectra of VraR_C with both R1‐DNA (Figure 5a) and R2‐DNA (Figure 5b). VraR_C titrated with R1‐DNA showed a greatly perturbed result for the residues Arg151, Ala169, Thr176, Lys177, Lys180, Ser184, Asp194, Ala198, and Phe203. The largest chemical shift changes were mapped on the VraR_C structure (Figure 5c). Of note, detected residues of CSPs with R1‐DNA agreed with the interactions found in the crystal structure of the VraR_C–DNA complex. However, VraR_C titrated with R2‐DNA showed less perturbed residues than R1‐DNA. In particular, residues Lys177 and Thr178 from the α9 helix were crucial in binding to the major groove of R1‐DNA but did not show any perturbation upon titrating with R2‐DNA. Perturbed residues from VraR_C titration with R2‐DNA mapped on VraR_C structure showed fewer residue involvement on binding to R2‐DNA (Figure 5d). A previous study showed that R1‐ and R2‐DNAs have two binding sites; however, R1‐DNA is a quasi‐palindromic sequence (₋₆₆AGACTAAAGTATGAACATCATT₋₄₉) with a single nucleotide separating two putative binding sites, whereas R2‐DNA has four additional nucleotides in between the first and second binding sites (₋₄₇GTTCCGGAGCCTATTCATATTGGTT₋₂₂). ²⁰ Possibly the low sequence similarity and distance between binding sites for R1‐ and R2‐DNA may result in the disparity in VraR_C titrations.

FIGURE 5.

Interaction of VraR_C with both R1‐ and R2‐DNAs inspected by NMR titrations. Plots of the normalized chemical shift perturbations of VraR_C after binding to R1‐DNA (a) and R2‐DNA (b). Gaps indicate the missing peaks from DNA titration. Perturbed residues with values >0.06 (dotted black line) mapped on the structure of VraR_C for both (c) R1‐DNA and (d) R2‐DNA titrations

2.5. The presence of Mg²⁺ ion enhances VraR_C –DNA binding affinity

The crystal structure of the VraR_C–DNA complex consists of Mg²⁺ ion bound to Asp194 in both protomers (Figure 6). Mg²⁺ ion was not added to VraR_C in any of the purification steps nor in complex formation with DNA; however, the crystallization condition contains 200 mM MgCl₂. The electron density available near Asp194 was filled with Mg²⁺, which is coordinated with five water molecules. Each Mg²⁺ ion binds to the OD2 of the Asp194 side chain in two subunits, with distances of 2.0 and 2.1 Å. The built structure with Mg²⁺ ion was assessed by using the CheckMyMetal server ³¹ , ³² to confirm the acceptable coordination interactions. To evaluate the importance of Mg²⁺ in binding affinity for VraR_C toward R1‐DNA, we performed BLI experiments in the presence and absence of Mg²⁺. The purified VraR_C was treated with 0.5 mM ETDA for chelating non‐specific metal bindings and finally buffer exchanged to 20 mM Bis‐Tris, 80 mM NaCl at pH 6.0. The binding affinity of VraR_C toward R1‐DNA was 3.71 μM in the absence of Mg²⁺ ions (Figure 7a). Of note, the addition of 10 mM MgCl₂ to VraR_C increased the binding affinity to 0.47 μM (Figure 7b).

FIGURE 6.

Mg²⁺ ions shown as green spheres binding to Asp194 of two subunits in the VraR_C–DNA complex. Subunits are shown in orange and blue. Each Mg²⁺ ion has a hydrogen bond interaction to the OD2 atom of Asp194 with a distance of 2 Å and 2.1 Å. Five water molecules coordinating the Mg²⁺ ions are in red spheres. The 2Fo‐Fc electron density map of Asp194, water molecules and Mg²⁺ ions was contoured at 1σ and shown as a blue mesh

FIGURE 7.

R1‐DNA binding affinity of VraR_C at different concentrations measured with BLI technology both in the absence (a) and presence (b) of 10 mM MgCl₂. All data were processed and analyzed by a double reference subtraction method by using Octet Data Analysis Software. The 2:1 binding model of association and dissociation function was used to determine K _d in the global fitting. (c) The electrostatic potential surface of VraR_C was calculated with the Pymol plugin APBS, showing red for the negative surface and blue for the positive charge surface. A slight negative charge appearing near the side chain of Asp194 was labeled. (d) Thermal denaturation curves of VraR_C–DNA complex in the presence (blue) and absence (red) of 10 mM MgCl₂ were monitored at 222 nm, revealing melting temperatures of 61.0 and 56.3°C, respectively

The electrostatic potential around the residues complementing DNA binding in the VraR_C dimer displayed a predominant positively charged surface (Figure S6). However, at the site of the minor groove interaction, the surface shows a partial negative charge from the carboxylic acid group of Asp194 (Figure 7c). The positively charged Mg²⁺ ion bound to Asp194 may cooperate with DNA binding. Also, we tested the thermal denaturation of the VraR_C–DNA complex in the absence and presence of Mg²⁺ ions. Although the denaturation curve became less cooperative with the addition of Mg²⁺ to the VraR_C–DNA complex, we observed a moderate increase in melting temperature from 56.3 to 61.0°C (Figure 7d). Apart from the two Mg²⁺ ions bound to Asp194, four Mg²⁺ ions were observed in the VraR_C–DNA complex structure, with two having no direct contact with the nucleotides and coordinated by six water molecules (Figure S7). The remaining two Mg²⁺ ions are bound to nucleotides C₄ and C_11′. Each Mg²⁺ ion was interconnected to each phosphate oxygen atom of two cystine nucleotides from nearby asymmetric units (Figure 8). Such a phenomenon might result from crystal contacts as well.

FIGURE 8.

Packing within the crystal of VraR_C–DNA complex is shown among three asymmetric units in pink, blue and yellow. Crystal contacts of DNAs were observed between two C₄–C₄ and C_11′–C_11′ nucleotides. The detailed Mg²⁺ ions interaction within two asymmetric units between DNAs are in boxes. Each Mg²⁺ ion (green spheres) binds with OP1 (asymmetric unit1) and OP2 (asymmetric unit2) atoms of C₄ nucleotides with 2.0 and 2.1 Å and are surrounded by water molecules (red spheres)

3. DISCUSSION

In S. aureus, VraR plays a substantial role in regulating several genes, which leads to a thick cell wall under stressful conditions. ¹⁴ , ²² Exploring VraR interaction behavior toward its cognate DNA might shed light on its action mechanism. Here, we present the crystal structure of the VraR_C–DNA complex consisting of two VraR_C molecules binding with R1‐DNA as a compact dimer. Structural alignments between the VraR_C–DNA complex and the crystal structures available from the same family of RRs, such as LiaR from Enterococcus faecalis, ³³ DosR from Mycobacterium tuberculosis ³⁴ and NarL from Escherichia coli, ³⁵ showed similar binding orientations, with RMSDs of 0.33, 0.91, and 0.98 Å, respectively.

The dimerization of active VraR with conformational flexibility in the linker region leverages its DBD to specific DNA recognition. ²⁸ A previous study showed that both inactive and active VraR interacts with R1‐DNA. ²⁰ Our AUC‐SV experimental analysis with Sedfit resulted in a molecular weight of ~64.5 kDa for active VraR with R1‐DNA, thus further validating the dimer formation in the complex. In contrast, the AUC‐SV results for inactive VraR and R1‐DNA indicated a partial complex formation. To understand the probable characteristics of inactive VraR with R1‐DNA, we overlaid the structure of VraR (PDB ID: 4GVP) and the VraR_C–DNA complex (Figure S8). Although all residues from the inactive VraR DBD fit well in the structural superimposition, the complex formation of inactive VraR and R1‐DNA was minimal on AUC‐SV analysis. The inaccessibility of the α10 helix in inactive VraR (because of closed conformation) might hamper the dimerization during the complex formation with R1‐DNA. Predominantly, the α10 helix is involved in dimerization from the available DBD‐DNA crystal structures in the NarL/FixJ family. ³³ , ³⁴ , ³⁵ Certainly, we cannot rule out that inactive VraR binding to DNA could promote an active‐like conformation or in tandem repeats, as observed in PhoP from the OmpR family RR. ³⁶

In RRs, Mg²⁺ ion plays a vital role in phosphorylation because it is susceptible to binding oxygen atoms mainly from the side‐chain carboxylates of conserved Asp residues typically in the RDs. Here, we report that two Mg²⁺ ions bound to each Asp194 in the VraR_C–DNA complex, and this is the first crystal structure showing the Mg²⁺ ions in DBD‐DNA structures from the NarL/FixJ family. Asp194 is a conserved residue in LiaR, and mutation of the corresponding Asp position to Asn significantly increased the binding affinity toward its promoter DNA. ³³ Also, the BLI assay showed an approximately eightfold increase in binding affinity for VraR_C toward R1‐DNA in the presence of MgCl₂. In addition, the crystal of VraR_C alone was grown with 200 mM MgCl₂ in crystal condition, but we found no indication of Mg²⁺ ion density map near Asp194. Thus, the Mg²⁺ ion might prefer binding to Asp194 only in the presence of DNA. Nevertheless, further validation of the Mg²⁺ ion role in VraR_C is needed.

4. CONCLUSIONS

Here, we describe the crystal structure of VraR_C from S. aureus in complex with its promoter DNA. Our crystal structure with biophysical analysis supports the binding of activated VraR to R1‐DNA in dimer formation. NMR titrations of VraR_C toward R1‐DNA also confirm the crucial residues in binding to R1‐DNA and agree with the crystal structure of the VraR_C–DNA complex. We also report the involvement of Mg²⁺ ion in the VraR_C–DNA complex, which was not present in the crystal structure of VraR_C despite adding Mg²⁺ ions during crystallization. Mg²⁺ ion might play a vital role in VraR_C, binding to Asp194 in the presence of DNA and increasing the binding affinity.

5. MATERIALS AND METHODS

5.1. Expression and purification of VraR and VraR_C

The gene encoding full‐length VraR (residues 1–209) or VraR_C (residues 138–209) was PCR‐amplified from the genomic DNA and cloned into the pET‐GB1 or pET29b (Novagen) vector with Nde I and Xho I restriction sites to obtain the recombinant protein with a C‐terminal His‐tag. All proteins were overexpressed in E. coli BL21(DE3). For labeled (¹⁵N, ¹⁵N/¹³C, and ¹⁵N/¹³C/²H) protein samples, the cells were grown in H₂O‐ or D₂O‐containing M9 minimal medium containing ¹⁵NH₄Cl (1 g/L) and 13C‐glucose (2 g/L) at 37°C. When OD600 readings reached 0.6, the cells were induced with 0.5 mM IPTG for VraR or 0.6 mM IPTG for VraR_C, then grown for an additional 4 hr. The cultured cells were pelleted down at 6000 rpm for 20 min. The cultured cells were lysed by using a microfluidizer in binding buffer (20 mM Tris–Hcl pH 8.0, 500 mM NaCl, 2% glycerol) and purified by nickel‐nitrilotriacetic acid (Ni‐NTA) affinity chromatography. Furthermore, the target protein was dialyzed and concentrated with buffer (20 mM Bis‐Tris pH 6.0, 80 mM NaCl, 2% glycerol) and purified with a size exclusion column (Superdex75 10/300). The purity of the recombinant proteins was verified on SDS‐PAGE. Protein concentration measurement was quantified by the BCA method according to the protocol provided by the manufacturer, using bovine serum albumin as a standard.

5.2. Double‐stranded DNA preparation and formation of complex with VraR

The single‐stranded DNAs of R1 and R2 from the vraSR operon were purchased from MDBio Inc. (Taiwan). For preparing double‐stranded DNA, equal amounts of two complementary deoxynucleotides were mixed, followed by heating at 95°C for 10 min and a slow cool‐down to room temperature. Phosphorylation of aspartate in RRs can be mimicked by using beryllofluoride. ³⁷ Leonard et al. showed the stable formation of VraR homodimerization upon treatment with BeF₃ ⁻²⁸. VraR was incubated with 5 mM BeF₃ ⁻ overnight at room temperature, and the activated protein was titrated with dsDNA in a 2:1 ratio. Crystallization of the VraR_C–DNA complex involved using a 2:1 ratio of two VraR_C molecules to one R1‐DNA duplex.

5.3. VraR_C with/without R1‐DNA (22‐bp) complex crystallization

Protein crystallization involved using the vapor diffusing method at 20°C. VraR_C formed crystals with 100 mM Tris–Hcl, pH 8.5, 16% PEG4000 (w/v), and 200 mM MgCl₂ with the sitting drop vapor diffusion technique. The crystallization conditions for VraR_C with R1‐DNA (22 bp) were screened by the hanging drop vapor diffusion technique. The crystallization drop of VraR_C or VraR_C–DNA at 7 mg/ml consisted of 1 μl with an equal volume of mother liquor. Crystals of VraR_C–DNA were grown for 4–7 days with optical conditions of 100 mM sodium cacodylate, pH 6.6, 10% PEG1500 (w/v), 5% PEG400 (w/v), and 200 mM MgCl₂. All crystals were cryoprotected with the addition of 15% to 25% glycerol in mother liquor and flash‐frozen in liquid nitrogen. X‐ray diffraction data were collected from beamline 05A at the National Synchrotron Radiation Research Center (Hsinchu, Taiwan). All diffraction data were indexed, integrated, and scaled by using HKL2000. ³⁸ The crystal structures were determined by molecular replacement in the PHENIX Phaser, ³⁹ with the VraR crystal structure (PDB ID: 4IF4) used as the search model. All initial models were manually rebuilt by using the program Coot ⁴⁰ and then refined in PHENIX. ⁴¹ Figures related to protein structures were generated by using Discovery Studio visualizer and PyMOL. ⁴²

5.4. CD spectroscopy

All CD spectra were collected by using a Chirascan‐plus qCD spectrometer (Applied Photophysics, Surrey, UK). Protein samples are prepared in 20 mM sodium phosphate, at pH 6.0, 7.0 and 8.0. The far‐UV spectra were acquired at 25°C with 15–20 μM protein samples in a 1‐mm path‐length cuvette. The signals from 195 to 260 nm were recorded three times with a scan rate of 20 nm/min and a bandwidth of 1 nm after subtracting the blank signals from a solvent. Equilibrium thermal‐denaturing experiments were obtained by measuring the change of CD signal at 222 nm from 25 to 95°C at a rate of 1°C/min with a 1°C interval.

5.5. Biolayer interferometry

We used the ForteBio Octet RED96 system, and the experimental conditions were performed at 25°C. 5′‐biotinylated 25‐bp R1‐DNA was prepared in assay buffer (20 mM Bis‐Tris and 80 mM NaCl at pH 6.0) and immobilized onto streptavidin biosensor tips. BLI signals were measured for the immobilized response, followed by the association of VraR_C at different concentrations (10, 5, 2.5, 1.25, 0 μM) and subsequent dissociation into blank assay buffer. Octet Data Analysis software was used for processing the collected data with a double reference subtraction method. The dissociation constant (K _d) was obtained by fitting with a 2:1 binding model of association and dissociation functions with the global fit.

5.6. NMR spectroscopy and data processing

The NMR sample of VraR_C (at 0.6 mM) was prepared in buffer containing 20 mM sodium phosphate, 50 mM NaCl and 1 mM NaN₃ at pH 7.0. The NMR spectra were acquired at 310 K on Bruker AVANCE 600, 800, and 850‐MHz spectrometers equipped with a z‐gradient TXI cryoprobe (Bruker, Karlsruhe, Germany). 3D triple‐resonance experiments of HNCO, HN(CA)CO and HNCACB spectra were performed with a ¹⁵N/¹³C/²H‐uniformly labeled protein sample in H₂O for backbone resonance assignments of VraR_C. Briefly, we used the mixture of 10% ¹³C‐glucose and 90% unlabeled glucose as the carbon source for the expression of the VraR sample and analysis of the distribution of ¹³C labels of the methyl groups in Leu and Val residues on a 2D ¹H, ¹³C HSQC spectrum. The weighted CSPs for backbone ¹⁵N and ¹H_N resonances were calculated by the equation Δδ = [((Δδ_HN)² + (Δδ_N/5)²)/2]^0.5 and for methyl ¹³C and ¹H resonances by the equation Δδ = [((Δδ_H)² + (Δδ_C × 0.3)²)/2]^0.5. For Leu and Val residues, the methyl CSP was the mean value of two methyl groups. All NMR spectra were processed by using Bruker TopSpin 4.0 or NMRPipe ⁴³ and analyzed by using NMRView. ⁴⁴

5.7. Analytical ultracentrifugation

Highly purified samples of inactive VraR (20 mM Bis‐Tris pH 6.0, 100 mM NaCl) and active VraR (20 mM Bis‐Tris pH 6.0, 100 mM NaCl, 5 mM BeF₃ ⁻, 5 mM MgCl₂) were used for AUC‐SV experiments on Beckman Coulter ProteomeLab XL‐1 ultracentrifuge. Approximately 20 μM inactive or active VraR mixed with R1‐DNA (30 bp) was loaded to An‐50TI rotor, and data collection was performed at 50,000 rpm and 280 nm under 20°C. SednTrep and SedFit were used for processing and analyzing the data.

5.8. Additional information

The atomic coordinates and structure factors of VraR_C, VraR_C‐DNA complex and VraR_N were deposited in the RCSB Protein Data Bank (PDB ID: 7VE4, 7VE5, and 7VE6). The backbone NMR chemical shifts of VraR_C were deposited in the Biological Magnetic Resonance Data Bank (accession no.: 51095).

CONFLICT OF INTEREST

The authors declare that there is no conflict of interest regarding the publication of this paper.

AUTHOR CONTRIBUTIONS

Jangam Vikram Kumar: expression, purification and crystallization of protein–DNA complex, determining the crystal structures, manuscript writing; Tien‐Sheng Tseng: crystallization and AUC‐SV experiments; Yuan‐Chao Lou: NMR backbone assignment; Shu‐Yi Wei: cloning, NMR CSP experiments; Tsung‐Han Wu: CD data analysis, Hao‐Cheng Tang: BLI data acquisition and analysis; Yi‐Chih Chiu: X‐ray diffraction data collection and processing; Chun‐Hua Hsu: validation and Funding; Chinpan Chen: conceptualization, project administration, funding, supervision and manuscript writing.

Supporting information

Table S1 Thermal stability of VraR_C, VraR, and BeF₃ ⁻ activated VraR in sodium phosphate buffer at the different pH values monitored at 222 nm by CD spectroscopy.

Table S2. DNA sequences used for binding and co‐crystallization with VraR_C. Potential binding nucleotide sites to the DNA binding domain identified from the previous study are underlined.

Table S3. Data collection and structural refinement statistics for VraR_C, VraR_C–DNA complex and VraR_N. Data were collected on a single crystal for each dataset.

Figure S1 The purified VraR_C and VraR on 15% SDS‐PAGE. Lane 1, unlabeled VraR_C; lane 2, ¹⁵N¹³C²H VraR_C; lane 3, VraR; lane 4, ⁵N¹³C²H VraR.

Figure S2 The AUC‐SV profiles showing the dimer formation of VraR with BeF₃ ⁻ activation. The c(S) distribution of inactive VraR is at 2.8 S and BeF₃ ⁻ activated VraR at 3.45 S.

Figure S3 (A) Far‐UV CD spectra for VraR_C, VraR and VraR activated by BeF₃ ⁻ in sodium phosphate buffer at pH 6 at 25°C. (B) The thermal unfolding experiments of VraR_C, inactive and active VraR are monitored at 222 nm reveal melting temperatures of 58.30, 48.82, and 52.17°C.

Figure S4 The crystal structure of VraR_C contains residues Glu143 to Gln209 and all α‐helices are labeled accordingly.

Figure S5 2D ¹H–¹⁵N HSQC spectra for VraR_C at pH 7.0 and 310 K. All backbone resonances were identified and labeled except for the amide resonances of Lys139 and Asn165. The side chain NH₂ resonances of Asn and Gln are connected by blue horizontal lines.

Figure S6 The electrostatic potential surface of two VraR_C subunits from the VraR_C–DNA complex having a predominantly positive charge favorable for DNA interactions. The electrostatic potential was calculated with Pymol plugin APBS, showing red for negative and blue for positive surface.

Figure S7 In the VraR_C–DNA complex, 4 Mg²⁺ ions (green spheres) are present near DNA. Two Mg²⁺ ions having no direct interactions with DNA are in black circles. Six water molecules coordinating each Mg²⁺ ion are in red spheres.

Figure S8 Structural overlay of inactive VraR (PDB ID: 4GVP) (yellow) toward a single subunit of VraR_C (blue) from the R1‐DNA complex. All 11 residues binding to R1‐DNA from the VraR_C–DNA complex are depicted as sticks on inactive VraR to show the probability of a binding pattern as the α9 helix fits into the major groove DNA.

Kumar JV, Tseng T‐S, Lou Y‐C, Wei S‐Y, Wu T‐H, Tang H‐C, et al. Structural insights into DNA binding domain of vancomycin‐resistance‐associated response regulator in complex with its promoter DNA from Staphylococcus aureus . Protein Science. 2022;31(5):e4286. 10.1002/pro.4286