Figure 1. A simplified view of the pap promoter region. The cooperative binding of Lrp to the proximal sites inhibits transcription of papBA genes, and GATC^dist is methylated in phase-OFF cells. In phase-ON cells, PapI interacts with Lrp and promotes its binding to the distal sites through an unknown mechanism. GATC^prox is open for methylation, and transcription can be turned on.

Figure 2. ¹H-¹⁵N HSQC spectrum of PapI labeled with the backbone assignments. The spectrum was collected at 25 C with a ¹⁵N, ¹³C-labeled sample in 50 mM potassium phosphate (pH 6.0), 150 mM KCl, 5 mM β-mercaptoethanol, 100 μM EDTA, 8% (v/v) ²H₂O, 0.02% (w/v) sodium azide. Assignments were made from Asn3 through His77 (except Pro48). Asterisks indicate residues aliased in the ¹⁵N dimension. Boxed peaks are from unassigned side-chains. Trp57 HN^ε1, located at 10.1 ppm along the ¹H dimension and 112.1 ppm along the ¹⁵N dimension, is not shown.

Figure 3. Reults of the structure calculation. (a) Superposition of the 25 lowest-energy structures. Their backbone rmsd was 0.40 Å over the structured parts of the protein (residues 2–11, 17–24, 28–40, 44–47, and 56–59). This Figure shows only the residues 1 through 73 (the C-terminal residue in the WT protein), and the eight extra residues at the C terminus of this construct are not shown. (b) The ribbon diagram of the averaged-minimized PapI structure (PDB ID 2HTJ). Its unstructured C-terminal tail (residues 63–81) is not shown. (c) The observed C^α chemical shifts of residues 2–77 were compared with their respective random-coil chemical shifts, and their differences were plotted. With respect to the random-coil values, consecutive downfield shifts (positive trends) indicate α-helices, and upfield shifts indicate β-strands (negative trends). (d) An NOE pattern indicative of antiparallel β-sheets was observed between V44–S47 and Y56–L59 patches. A similar but weaker interaction was detected between Y56-W57 and G15-K16 (broken arrows).

Figure 4. Backbone ¹⁵N relaxation data. (a) The order parameters (S²), (b) effective correlation time (τ_e), and (c) {¹H}¹⁵N-NOE are shown, along with (d) the averaged rmsd calculated from the backbone non-hydrogen atoms of the 25 lowest-energy structures shown in Figure 3(a). In (d), rmsd for residues 78–81, for which dynamics data are unavailable, is not shown.

Figure 5. The DNA-binding experiments. A deuterated, ¹⁵N-labeled PapI sample was titrated with the pap 2, pap 5, and randomized pap 5 duplex DNA. (a) A part of the ¹H-¹⁵N correlation spectrum of PapI corrected at different pap 5 DNA concentrations are overlaid. Only the 0 μM (blue), 35 μM (cyan), 90 μM (yellow), and 660 μM (red) DNA data points are shown for clarity. T17 resonance was broadened to the extent that it cannot be seen at this contour level. (b) Titration curves were constructed for several representative residues, including Q45 ¹H shown here. A model assuming a 1:1 stoichiometry did not fit the data (red line). The model involving 2-PapI-to1-DNA improved the fit (green line). This was a consistent trend shared by all residues examined. (c) The chemical shift difference between the DNA-free state and saturated state was plotted across the PapI sequence. Data for P48 and T55 are not available. (d) The total chemical shift perturbation in the presence of DNA shown in (c) was color-coded and plotted onto the PapI structure in two different orientations. Red indicates large chemical shift differences. Residues in blue had smaller shift changes. (e) The electrostatic potential map of PapI (front and back). Basic residues are colored blue, and acidic residues are colored red. In (d) and (e), only residues 1–62 are shown.

Figure 6. A model for PapI-DNA interaction. (a) PapI residues 1–62 are superimposed with the DNA-binding domain of Staphylococcus aureus MecI complexed with the mec promoter (PDB ID 1SAX) to 1.3 Å rmsd. (b) The MecI dimer was removed and the figure is rotated slightly relative to (a) for clarity.

Figure 7. Sequence alignment of selected FaeA-like family members (PF04703). Black residues are highly conserved. Gray positions are occupied by functionally similar residues. The PapI secondary structure (H, α-helix; S, β-strand) is shown underneath. Abbreviations: S, Salmonella; E, Escherichia; P, Pyrobaculum; hyp, hypothetical.

Statistical evaluation of the structure calculation

Numbers in parentheses indicate the total numbers of constraints in the categories.

Structural neighbors of PapI (selected entries from the VAST search results)

Search for the structural neighbors of PapI was done using its first 62 residues or only the 39 residues in the helices and strands (results in square brackets). S., Staphylococcus, M., Methanococcus, A., Archaeoglobus, P., Pseudomonas; NA, not available.