Crystal structure of Enterococcus faecalis SlyA-like transcriptional factor Structure of SlyA reveals winged-helix DNA-binding motif

The crystal structure of a SlyA transcriptional regulator at 1.6 ¯ resolution is presented and structural relationships between members of the MarR/SlyA family are discussed. The SlyA family, which includes SlyA, Rap, Hor and RovA proteins, is widely distributed in bacterial and archaeal genomes. Current evidence suggests that SlyA-like factors act as repressors, activators, and modulators of gene transcription. These proteins have been shown to upregulate expression of molecular chaperones, acid-resistance proteins, and cytolysin, and to downregulate several biosynthetic enzymes. The crystal structure of SlyA from Enterococcus faecalis , determined as a part of an ongoing structural initiative (www.mcsg.anl.gov), revealed the same winged-helix DNA-binding motif that was recently found in the MarR repressor from Escherichia coli and the MexR repressor from Pseudomonas aeruginosa , a sequence homologue of MarR. Phylogenetic analysis of the MarR/SlyA family suggests that Sly is placed between the SlyA and MarR subfamilies and shows significant sequence similarity to members of both subfamilies.

The slyA gene is implicated in virulence and environmental adaptation of S. typhimurium. It has been shown to upregulate expression of molecular chaperones, acid-resistance proteins, and cytolysin, and to downregulate several biosynthetic enzymes [11]. It has also been suggested that regulation of gene expression by SlyA is crucial for intracellular survival and/or replication of both enteroinvasive E. coli and Salmonella serovar Typhimurium in phagocytic host cells.
RovA, whose amino acid sequence is 77% identical to the S. typhimurium SlyA, plays a role in regulation of the invasion of mammalian cells by Yersinia and mediates regulation of invasin in response to environmental signals [12]. Invasin is a primary factor that allows efficient internalization of Y. pseudotuberculosis into eukaryotic cells. All the current evidence shows that SlyA-like factors act as repressors, activators, or modulators of gene transcription.
The SlyA protein from a pathogen Enterococcus faecalis (SlyA-Ef) was selected for this study as a part of an ongoing structural initiative (www.mcsg.anl.gov). Specific biochemical function, the signal molecule and the DNA-binding properties of this protein are unknown. Phylogenetic coli strain BL21[DE3] carrying the pMAGIC vector was induced with isopropyl-β-Dthiogalactoside. Cells were harvested after 4 hours of culture at 37°C; suspended in 50 mM phosphate buffer pH 8.0, 300 mM NaCl, 10 mM imidazol, 10 mM β-mercaptoethanol, and 10% glycerol; and lysed by sonication. The fusion protein was purified by affinity chromatography using Ni-NTA Superflow resin (Qiagen). The His-tag was removed by digestion with rTEV and the resulting protein was purified by affinity chromatography using Ni-NTA Superflow resin (Qiagen). In this design three amino acid residues were added at the N-terminus of SlyA. The cleaved protein was further purified on an SP Sepharose Fast Flow column (Pharmacia) using 0.5 and 1.0 M NaCl two-step elution and concentrated with simultaneous buffer exchange using Centriplus-3 (Amicon) (3-kDa cutoff). A 2 mM protein stock solution in 10 mM Tris-HCl pH 7.4, 20 mM NaCl, and 1 mM DTT was used for crystallization. Selenomethionine-labeled SlyA protein was prepared using methionine biosynthesis inhibition method [13]. Crystals of native SlyA protein and its SeMet derivative diffracted to 1.8 Å and 2.0 Å, respectively. The space group was P2 1 2 1 2 1 with cell dimensions of a = 43.512, b = 48.429, c = 95.911. MAD data were collected to 2.0 Å resolution from a single crystal containing SeMetlabeled protein at three different x-ray wavelengths near the Se edge. Inverse beam strategy was used. The absorption edge was determined by fluorescent scan of the crystal as described previously [13]. The data were processed using HKL2000 suite [14]. Crystal characteristics and data collection statistics are presented in Table 1.
The structure of SlyA was determined using the MAD phasing method. All four Se sites were found in the asymmetric unit. MAD phases were calculated using the CNS suite [15] phases and improved using the density modification (DM) method as implemented by the CNS suite.
Electron density maps were high quality and allowed autotracing of the amino acid chains using 7 The DNA-binding domain is formed by the central part of the polypeptide and includes α-helices H2, H3, H4, and H5, and a β-hairpin (residues 73-91). There is virtually no interaction between DNA-binding domains in the SlyA dimer. This domain contains a DNA-binding motif described previously as a winged-helix. In SlyA-Ef, it is composed of H5 and the 19-residue β-hairpin as a wing. This architecture closely resembles the WH motif of the well-established major-groove DNA-binding domain found in a number of transcriptional regulators [6,[19][20][21][22].
The WH motif in SlyA-Ef is preceeded by two short α-helices, H3 and H4, with a short β-strand and a small loop between them. The motif is followed by H6, which is a part of the dimerization domain. Two residues conserved in the SlyA family, Thr91 and Gly94, are located at the junction between the wing and H6 (Fig. 1). These residues are also conserved in the MarR family and may be important for the positioning of the DNA-binding domain on its DNA target. The MexR structure shows three different dimers with varied spacing between their WH motifs and high degrees of conformational flexibility [7]. The largest space between Cα positions of Arg73/Arg73' was 29.9 Å. In the SlyA-Ef structure, the equivalent spacing (Arg67/Arg67') is shorter (24.63 Å). Analysis of the MarR structure using a program available at the PITA server to generate biological dimers [23] indicates that the spacing between the Cα positions of Arg77/Arg 77' is 28.1 Å. These key arginine residues exhibit different arrangements in the three structures. In MexR, they are positioned on the DNA-binding surface and are expected to point toward the bases in the DNA major groove. In MarR and SlyA, the arginine side chains are also at the DNA binding surface but are expected to point away from the bases and the major groove.
These differences suggest that the mode of binding of these proteins to DNA is not identical.
Moreover, these data suggest that MarR/SlyA proteins may span a narrow minor groove and may recognize a rather short palindromic DNA sequence (14-16 bp).
The high-degree of flexibility of DNA-binding domains displayed in different crystals provides indirect evidence of the ability of this fold to adapt in order to recognize various DNA targets.
Analysis of the crystal structure of the complex bound to DNA will test this hypothesis. It is also evident that the C-terminus of SlyA is in position to facilitate an interaction with DNA. The Cterminus of SlyA and its homologues contains several conserved positively charged residues that could make contacts with phosphate groups or bases.
In the SlyA-Ef dimer, the WH motifs are positioned to interact with a palindromic recognition sequence, with specific contacts expected to occur predominantly between H5 and the wing, and the major groove of DNA. The WH motif in SlyA-Ef is strongly positively charged due to the presence of five arginines and four lysines; however, these residues are not conserved between SlyA and MaR families, suggesting that different DNA targets or different sets of residues in the same DNA target are recognized.
A cavity between DNA-binding domains in the SlyA-Ef dimer and extending to the dimerization domain could serve as a binding site for a signal molecule. In MexR, a C-terminal tail of the C monomer is inserted into this cavity [7]. The residues lining the cavity are rather hydrophobic, but are not conserved, suggesting that these proteins respond to different signals. Interestingly, binding of a single ligand molecule to the dimer is expected to perturb the two-fold symmetry, allowing the molecule to detect differences in the DNA target sequence. A signal molecule could also control spacing between DNA-binding domains by arranging them differently. Another key difference between SlyA and MarR is that the entrance to the cavity in MarR is closed off by two salt bridges between Asp67 and Arg73. These interactions are not present in the SlyA or MexR proteins. Recent evidence suggests that SlyA-dependent regulation my involve other protein cofactors, such as cAMP receptor or nitrate reduction regulator [24], which may involve direct protein-protein interactions.
Overall, the SlyA structure is very similar to the structures of MarR and MexR, showing that the MarR/SlyA family shares a common fold and similar DNA-binding properties despite their low amino acid sequence similarity. However, these proteins show significant local divergence in sequence and structure that apparently allows them to respond to different signal molecules and to bind to diverse DNA targets by using alternative modes of action. The MarR/SlyA structure is an example of a highly adaptable protein fold and provides a plausible explanation for its wide use in transcription regulation in many different organisms.