The crystal structure of the effector‐binding domain of the trehalose repressor TreR from Bacillus subtilis 168 reveals a unique quarternary assembly

The trehalose utilization system in Bacillus subtilis 168, consisting of the two enzymes trehalose permease and phosphate-α-1,1-glucosidase, is negatively regulated at the transcriptional level by the trehalose repressor (TreR).1 TreR belongs to the HutC subfamily of the GntR superfamily2 of bacterial regulatory proteins and contains an N-terminal helix-turn-helix DNA-binding domain and a C-terminal effector-binding domain. When no trehalose is present in growth media, TreR interacts with two target sequences (operators) in the tre operon and blocks transcription. By contrast, the availablity of trehalose induces expression of genes in the tre operon. In in vitro experiments, trehalose-6-phoshate was identified as the molecular inducer that blocks TreR binding to its operator.1, 3 Here we present the crystal structure of the TreR effector-binding domain using diffraction data to 2.5 Å resolution. The three-dimensional structure reveals a unique oligomeric assembly that is likely important for TreR function. The TreR C-terminal effector-binding domain (C-TreR, residues 90–238) structure was determined by single-anomalous dispersion (SAD) using crystals of the selenomethionyl-substituted protein and was refined using data to 2.5 Å resolution. Our crystal form contains one molecule in asymmetric unit and 69% solvent. The final crystallographic model comprises residues 95–238, with the five N-terminal residues being disordered. The C-TreR, similar to other HutC regulators,4 exhibits the UbiC fold found in chorismate lyase from E. coli5 and is characterized by a central six-stranded antiparallel β-sheet and three short helixes (Fig. 1). The 10 C-terminal residues (227–238) form an extended strand that packs against β-strand S4 of a neighboring molecule related by crystallographic symmetry and forms a tight interface (Fig. 1). This arrangement is consistent with the oligomeric state of the protein in solution. Dimerization mediated by an effector domain is a general feature of characterized GntR family members,2 and TreR was observed to exist exclusively as a dimer in solution.3 Dimer interfaces similar to the one in the C-TreR structure can be found in other structures of effector-binding domains of GntR-like regulators: PhnF from E. coli (PDB code 2FA14) and YurK from B. subtilis (PDB code 2IKK). Ribbon diagram of the C-TreR dimer. α-helices and β-sheets are numbered consecutively and colored red and yellow, respectively. In addition to the dimeric interface, the association of two dimers into a tetrameric assembly can be observed in our crystal structure (Fig. 2). The overall solvent-accessible surface area buried upon tetramer formation is 9255 Å.2 It is thus possible that TreR is able to adopt a tetrameric quaternary structure in solution, a notion that is supported by an analysis made with PISA.6 The C-TreR tetramer. One monomer is represented with its solvent-accessible surface showing the distribution of conserved residues among 200 sequences homologous to TreR. The surface is colored in rainbow colors, ranging from blue representing sequence identity of 10% or lower to red representing sequence identity of over 90%. The sulfate ion is shown as a stick model in white. A schematic of the tetrameric assembly is shown in the upper right corner. Monomers of the two dimers are colored blue/red and gray/magenta, respectively. In addition to its dimeric partner, each monomer within the tetramer interacts with the two monomers of the neighboring dimer (Fig. 2). The interaction with one of these monomers is through helix H3 (residues 162–170), and the interaction with the second monomer is mediated by C-terminal residues. The C-termini interact with a surface cavity formed by residues that are highly conserved in GntR regulators (Fig. 2). This cavity probably represents the TreR effector-binding site. A sulfate ion is bound to this cavity in our crystal structure, possibly mimicking the position of the phosphoryl group of trehalose-6-phosphate effector molecule. This sulfate ion binds to a region that is equivalent to the product-binding region in UbiC chorismate lyase.5 If the conserved surface region indeed represents the effector-binding site, modulation of TreR activity could be achieved by the C-terminal interactions competing with the binding of the effector molecule. No similar quarternary arrangement was found in any of the available structures of related regulators. Among the GntR regulators with known operator sequences, TreR is the only regulator that recognizes two operators, located close to each other (32 bp apart) on the DNA double-helix.2 A model of cooperative binding of two dimers of TreR to two operator sequences has been proposed based on in vitro DNA binding studies.3 Binding of trehalose-6-phosphate abolishes TreR oligomerization and thus the ability to bind the tre operators.3 The C-Tre crystal structure did not allow us to make inferences about the orientation of the DNA-binding domain in full-length TreR; however, the tetrameric assembly of the effector-binding domain with its unique interactions between the C-terminal regions of one dimer and the effector-binding sites of the neighboring dimer are in agreement with the proposed model of TreR modulation. The DNA sequence of the effector-binding domain of Bacillus subtilis TreR (residues 90–238) was inserted into the pMCSG7 expression vector7 and overexpressed in E. coli BL21 (DE3). For the production of the selenomethionyl-substituted variant, cells were grown using enriched M9 medium supplemented with selenomethionine.8 C-TreR was purified as described previously.9 Initial protein crystals were obtained by vapor diffusion in sitting drop mode by using solutions from the Index crystallization screen (Hampton Research) carried out using a HoneyBee crystallization workstation (Cartesian Technologies). Further optimization was done manually in the hanging drop mode. Improved crystals were obtained by mixing 1 μL of protein (21 mg/mL in 10 mM HEPES pH 7.5, 125 mM NaCl) with 1 μL of reservoir solution (1.26M ammonium sulfate, 0.1M MES pH 6.0, 0.56 mM CYMAL-5) and equilibration at 20°C over 1 mL reservoir solution for 5–6 weeks. For cryoprotection, crystals were transferred into reservoir solution supplemented with 20% (v/v) glycerol and flash-cooled in liquid nitrogen. Diffraction data were collected at 100 K at beamline 19-ID of the Structural Biology Center at the Advanced Photon Source, Argonne National Laboratory, using X-rays with a wavelength of 0.979 Å. The diffraction data were processed using the HKL2000 suite of programs.10 The C-TreR crystals exhibited anisotropic diffraction to a minimum Bragg spacing (dmin) of about 2.5 Å in one direction (along c) and a dmin of about 2.8 Å in the other directions. Crystal parameters and data collection statistics are summarized in Table I. The C-TreR structure was determined by SAD phasing using HKL3000.11 An initial model was automatically constructed using the program ARP/wARP.12 The electron density was well defined and connected for the entire protein, with the exception of the five N-terminal residues, which could not be located in the electron density. Refinement was carried out using REFMAC 5.2,13, 14 and manual rebuilding was done using COOT.15 Overall anisotropic scale factors calculated during refinement illustrate the crystal diffraction anisotropy (Table I). The phasing and refinement statistics are summarized in Table I. The stereochemistry of the structure was evaluated with PROCHECK.16 Figures were prepared with PyMOL.17 Atomic coordinates and experimental structure factors of C-TreR have been deposited with the PDB and are accessible under the code 2OGG. The authors thank Min Zhou and other members of the Structural Biology Center at Argonne National Laboratory for their help with protein purification and conducting data collection at 19-ID and 19-BM beamlines.