Studies of IscR reveal a unique mechanism for metal-dependent regulation of DNA binding specificity

Abstract

IscR from Escherichia coli is an unusual metalloregulator in that both apo and iron sulfur (Fe-S)-IscR regulate transcription and exhibit different DNA binding specificities. Here, we report structural and biochemical studies of IscR suggesting that remodeling of the protein-DNA interface upon Fe-S ligation broadens the DNA binding specificity of IscR from binding the type 2 motif only to both type 1 and type 2 motifs. Analysis of an apo-IscR variant with relaxed target-site discrimination identified a key residue in wild-type apo-IscR that, we propose, makes unfavorable interactions with a type 1 motif. Upon Fe-S binding, these interactions are apparently removed, thereby allowing holo-IscR to bind both type 1 and type 2 motifs. These data suggest a unique mechanism of ligand-mediated DNA site recognition, whereby metallocluster ligation relocates a protein-specificity determinant to expand DNA target-site selection, allowing a broader transcriptomic response by holo-IscR.

Figure 1

Overall Structure of IscR bound to the hya promoter. (a)The IscR dimer is shown as a ribbon representation, with the monomeric subunits colored purple and cyan and the DNA rendered as a stick model. The positions of the ligating residues in IscR are shown as large spheres with Ala92, Ala98, and Ala104 in purple and His107 in pink. This cluster binding region is in close proximity of to the wHTH DNA binding domain of the adjacent monomer. The −35 hexamer of the hya promoter is shown in red. Electron density is not apparent for residues beyond residue numbers 136 and 144 in the two IscR subunits for the subunits shown in purple and cyan, respectively. (b) Cartoon representation of an IscR monomer. The dashed line indicates residues for which electron density was missing. (c) Superposition of B-form DNA (red) and the IscR bound hya promoter (blue).

Figure 2

IscR-hya DNA Interactions. (a) Schematic representation of contacts between IscR and hya DNA. Contacts from each monomeric subunit of IscR are shown in cyan and purple, respectively. (b, c) Residues involved in the interaction between the recognition helix of each monomer of IscR and the major groove of DNA. Yellow lines indicate hydrogen bonds. (d) View of the interaction between the wing of IscR and the minor groove of DNA.

Figure 3

IscR binding to the hya promoter involves only minor conformational changes. (a, b) Superposition of the non-DNA bound IscR structure with that of IscR bound to the hya promoter. The recognition helix of one monomer from each structure is shown as cylinders. (c) A close-up view showing the rotameric shift of the Glu43 side chain upon DNA binding.

Figure 4

Elimination of the Glu43 carboxyl group of apo-IscR broadens the DNA binding specificity to include type 1 sites. Binding isotherms for wild-type [2Fe-2S]-IscR (filled circles), wild-type apo-IscR (open circles), [2Fe-2S]-IscR-E43A (filled triangles), or apo-IscR-E43A (open triangles) with a DNA containing the iscRB site. Error bars represent the standard deviation of at least three separate assays.

Figure 5

Apo-IscR does not bind a type 1 site containing symmetrical cytosines at positions 7 and 8 of each half site. Binding isotherms of [2Fe-2S]-IscR (filled circles) and apo-IscR (open circles) with the iscRBcc site (5'-AAATAGCCGACCATTTTACTCGGGAATGTC-3', mutated dinucleotide is bolded, symmetrical positions 7 and 8 are underlined). Error bars represent the standard deviation of at least three separate assays.

Figure 6

Model for IscR discrimination between type 1 and type 2 motifs. Interactions of apo and [2Fe-2S] IscR at type 1 and type 2 sites are shown. The side chains of Glu43 (black), Ser40 (gray), and Gln44 (gray) of the recognition helix (RH) make contacts to the DNA major groove and Arg59 (gray) of the wing (W) contacts the minor groove. A black diamond on IscR indicates the presence of a [2Fe-2S] cluster. The key nucleotides in the hya type 2 site or conserved positions in type 1 sites are represented in bold by A (adenine), T (thymine), C (cytosine), or G (guanine). (a) The contacts between apo-IscR and the type 2 site. Glu43 of apo-IscR interacts with a C₇C₈ dinucleotide in one half site and C_7'A_8' the other half site. Additional contacts are made by Arg59 with the minor groove, Ser40 with an A or G, and Gln44 with T. (b) The contacts between [2Fe-2S]-IscR and the type 2 site are the same for apo-IscR, described in panel a. (c) The lack of interaction of apo-IscR with the type 1 site is indicated by the starburst. (d) The contacts between [2Fe-2S]-IscR and the type 1 site. Displacement of the Glu43 (black) side chain from the major groove is indicated by the outward-pointing arrows. The interaction of Arg59 with the minor groove is the same as in panels a and b. Question marks represent uncertainty about the site of the interaction of Ser40 and Gln44 with conserved nucleotides in the type 1 site.