The emergence of bacterial resistance to multiple drugs poses a serious and growing health concern. Understanding and deciphering these multiple drug-resistance mechanisms are important endeavors. The Escherichia coli transcriptional regulator MarR is believed to be a key factor in regulating marRAB operon, which is responsible for the mar phenotype that is resistant to a wide variety of structurally different and medically important antibiotics.1–3
The crystal structure of a salicylated-bound E. coli MarR was recently determined to a resolution of 2.3 Å.4 However, the free MarR crystals grown without salicylate were unstable and badly disordered to preclude structural determination.4 We have successfully crystallized a putative ligand-free MarR family protein (XC1739) from a plant pathogen Xanthomonas campestris and determined its structure to a resolution of 1.8 Å. It shares a 32% identity (55.6% similarity) with the E. coli MarR protein.4 A promoter sequence of 24 bases upstream of the XC1739 gene was also identified from the EMSA (Electrophoretic Mobility Shift Assay) and DNase I footprinting methods (Yang et al., unpublished result). These data, along with the finding that the structure of XC1739 is structurally similar to that of the E. coli MarR, suggest that XC1739 is very likely a transcriptional regulator belonging to the MarR family.
Materials and Methods.
The data on the gene cloning, protein expression, purification, crystallization, and diffraction for native XC1739 have been previously reported.5 Se-Met labeled XC1739 was produced using a nonauxotroph E. coli strain BL21(DE3) as host in the absence of methionine but with ample amounts of Se-met (100 mg/liter). The induction was conducted at 37°C for 4 h by the addition of 0.5 mM IPTG in M9 medium consisting of 1 g of NH4Cl, 3 g of KH2PO4, and 6 g of Na2HPO4 supplemented with 20% (w/v) of glucose, 0.3% (w/v) of MgSO4, and 10 mg of FeSO4. Purification and crystallization of the Se-Met-labeled XC1739 were performed using the protocols as established for the native protein.5
Crystals of Se-Met-labeled XC1739 were stabilized in mother liquor containing 22% PEG 2K MME (polyethylene glycol monomethyl ether) plus 20% glycerol before freezing in liquid nitrogen. Diffraction data sets were collected from a single Se-Met-labeled crystal using the SPring-8 beamline SP12B2 with an ADSC Quantum 4 CCD detector at 100 K. The diffraction data were processed using HKL2000,6 and indexed according to the same crystal orientation matrix. The structure of XC1739 was determined to 1.8-Å resolution using SAD method. The refinement of selenium atom positions, phase calculation, and density modification were performed using the program SOLVE/RESOLVE.7 The model was manually built using the XtalView/Xfit package.8 CNS9 was then used for refinement to a final Rcryst of 21.5% and Rfree of 25.3%, respectively. The unit cell parameters of Se-Met-labeled XC1739 crystals are identical to those of native protein and were summarized in Table I. The coordinates of XC1739 have been deposited (PDB, 2FA5).
The XC1739 crystals diffract to a resolution of at least 1.8 Å. Careful examination of the XC1739 electron density map allows us to trace amino acid residues from H12 to A162 unambiguously, except for the wing residues from T97 to S105 in the DNA-binding domain, which remain invisible. The N-terminal residues from H12 to L22 do not adopt any secondary structure, yet were clearly defined. Several hundred (313) water molecules and two chloride ions were also detected, and no other ligands with electron density above two standard deviations could be found.
The fold of XC1739 is similar to those of other MarR family members and consists of six α-helices and three β-strands with a α1-α2-β1-α3-α4-β2-β3-α5-α6 topology. We also observed that all highly conserved residues of XC 1739 were mapped to the similar positions of E. coli MarR throughout the dimerization domain, DNA-binding domain, and salicylate-binding residues [see Fig. 1(a)].4 The above arguments, along with the comparable three-dimensional structure described below [Fig. 1(b)] and the rather electropositive surface potential in the putative DNA binding regions [Fig. 1(c)], indicate that XC1739 is very likely a member of MarR family.
The XC1739 structure is well determined with all backbone torsional angles in the allowed regions (98% in the most favored regions, and 2% in the additional allowed regions) in the Ramachandran plot. The overall architecture of XC1739 structure is similar to that of E. coli MarR, comprising two monomers stabilized by the dimerization domains and containing two possible DNA-binding domains [Fig. 1(b)]. The N-terminal α1 and C-terminal α6 helices interact strongly through hydrophobic forces to stabilize the dimeric structure. Interestingly, a short extended π-helix was observed in the C-terminal end of α6 helix, causing an expansion of the helix diameter by approximately twofold in that region. The rarely observed π-helix was confirmed by the H-bonding formation between the N-H group of residue i and the carbonyl group of residue i + 5 (R149–L155 and L150–L155).10 It is also interesting to note that the charge distribution for XC1739 is, similar to E. coli MarR, rather uneven, with the dimerization domain enriched in acidic residues, yet the DNA-binding domain in basic residues, as shown in the electrostatic surface plot in Figure 1(c).
Although the overall architecture of XC1739 is similar to that of E. coli MarR,4 the putative DNA-binding domain of XC1739 is significantly shifted relative to that of the liganded E. coli MarR, as shown in Figure 1(d). When the two MarR proteins were superimposed using the Cα atoms of the left monomers, the right monomers exhibit significant shifting, especially for the α4 helices [Fig. 1(d)]. The only salt bridges detected between Asp67 and Arg73′ and the reciprocal pair in the salicylate-binding E. coli MarR are found disrupted in the ligand-free XC1739 (larger than 9 Å for XC1739 Asp76 and Arg82′). This may allow the DNA-binding lobes of ligand-free XC1739 to act independently for binding to XC1739 marO. Because there are multiple modes of DNA binding within the winged-helix family of DNA-binding proteins,11, 12 the manner in which XC1739 binds to DNA remains to be determined. This is currently in progress.
A structural homology search by the DALI program13 returns with two other protein structures of the MarR superfamily with good matching scores. The first one is a MexR responsible for the repression of MexAB–OprM operon that encodes a multidrug efflux system in Pseudomonas aeruginosa, resulting in increased resistance to multiple antimicrobials.14 The second one is an SlyA protein, which has been shown to upregulate the expression of molecular chaperones, acid-resistance proteins, cytolysin, etc., in Enterococcus faecalis.15 Besides, the coordinates of three other MarR family proteins have also been deposited in PDB (1s3j, 1a61, and 1z7u) recently. They also adopt similar triangular architecture despite the very low sequence identities (∼20%) (data not shown). Because the relative distances between their putative DNA-binding lobes vary a lot, the conformations of MarR family proteins are thus flexible and may adopt different binding modes with their cognate DNA promoters.
Recently, a paper describing the complex crystal structure of the ohrA operator with OhrR, an organic hydroperoxide-resistance transcriptional regulator belonging to the member of MarR family, was published.16 Detailed interactions between the OhrR DNA binding motifs, that is, an extended wing motif, an HTH motif, and an HH motif, and the ohrA operator with a palindromic 5′-ATTGTATACAAT-3′ sequence in the center, were described. Considerable shifting of the DNA binding domains in OhrR was also observed compared with the apo OhrR. However, the XC1739 binding sequence detected using EMSA and DNase I footprinting contains only a three base palindromic sequence at the 5′- and 3′-end of a DNA 24-mer (unpublished data). It is thus interesting to see how different transcriptional regulators in a same family adapt to recognize their DNA partners presumably through different behaviors. The cocrystal study between the XC1739 and its cognate DNA is currently underway.
We thank the Core Facilities for Protein Production in the Academia Sinica, Taiwan, for providing us the original vectors used in this study,17 and the Core Facilities for Protein X-ray Crystallography in the Academia Sinica, Taiwan, and the National Synchrotron Radiation Research Center, Taiwan, for the assistance of X-ray data collection. This work is also supported by the grant from the Department of Health, ROC (DOH94-TD-G-111-041).