A, Schematic of the PR-DBD648 construct. Alpha helices are boxed. Direct dimer contacts (L602, C603, A604, R606, D608, 1610, and R615) are indicated by solid dots, and the water-mediated dimer contacts (K617 and N618) are marked with open dots. DNA backbone interactions are marked according to whether they fall within or outside of the consensus PREs with solid (H578, Y579, R616, K617, and R623) and open squares (C577 and R617), respectively. Base-specific interactions are denoted with closed stars (K588, V589, R593). The CTE-DNA interaction (R637) is depicted with an open star. Water-mediated DNA contacts are marked with triangles (D578, C577, Y579, G580, S586, and K617). N-terminal residues that are cloning artifacts are in lower case (g and s). Residues that are disordered in the crystal (g, s, L562, K640, K641, F642, N643, K644, V645, R646, V647, and V648) are marked with a dotted line. B, The DNA fragment named PRE_2C7A used in the crystals. The consensus PREs are separated by 3N spacer nucleotides that are derived from mouse mammary tumor virus. Consensus inverted repeats are highlighted in gray, and PRE numbering is indicated below the sequence. A and B denote the subunit of PR-DBD bound to the hexanucleotide element.

A, Interactions between CTE residues R637 and K638 and the DNA. Red lines indicate predicted hydrogen bonds with the minor groove. Carbon, oxygen, and nitrogen atoms are in yellow, red, and blue, respectively. The DNA is colored gray and CTE interacting nucleotides are labeled. Gua8 is boxed to indicate that it is part of the adjacent asymmetric unit. The electron density is a CCP4 FWT map with the contour set at 0.4 σ. B and C, Quantitative EMSA with the palindromic PRE (B) and half-site PRE (C). Increasing concentrations of wild type PR DBD648, PR DBD R637A, and PR DBDR637A/K638A were incubated with a single limiting concentration of [³²P] labeled PRE (0.6 nM). The gel insets are representative experiments with DBD648 wild type, R637A, and R637K638A and the resulting binding curves represent mean values (±SEM) from replicate experiments (see Materials and Methods for details). D, Sequence alignment of steroid receptor and ERR2 CTEs. Occurrences of the well-conserved RK sequence in steroid receptor CTEs are indicated by bold print. PR and ERα XGGR sequence (aa 634–637 642–645) that closely resembles the VLRGGR (aa 642–646) sequence found in the orphan nuclear receptor subclass III are underlined. The ERR2 Tyr (aa 649), and potential residues for PR (F639, F641, V644, V647, and V648), which may make hydrophobic interactions with the core DBD, are marked with stars. Ribbon diagrams show monomers of PR and ERR2 bound respectively to PRE half-site and an ERE half-site (9). The arginines of the VXGGR sequences, which interact at comparable positions in the minor groove, are circled in black. Helix 1, 2, and 2′ are marked, the core DBD is colored blue and the CTE is colored red. Amino acid numbering is indicated for PR.

A, Graph of minor groove width (measured by the program CURVES) vs. nucleotide position of the DNA response elements. The protein names of the oligonucleotide used in the text, and the PDB IDs (R4R, 1R4O, 1HCQ, 1R4I) are given for each complex along with the sequence of the DNA fragment from steroid receptor DBD-DNA crystal structures. HREs are indicated by gray boxes. B, Analysis of the sequences selected by PR-DBD-648 by MEME. The sequence of the inverted repeat PREs and the preferred sequences in the selection are shown on the x-axis, and the information content of the alignment for each position is shown on the y-axis.

The ribbon diagram of the structure shows the PR-DBD in rainbow coloring from blue at the N terminus to red at the C terminus. Zinc ions and the DNA are colored gray. DBD subunits are labeled A and B, respectively. Helix 1 (H1), helix 2 (H2), and helix 2′ (H2′) are indicated on each DBD subunit. The DNA ladder diagram details the contacts made between the DBD subunit A and one hexanucleotide-half-site of the PRE. The three-nucleotide spacer between the inverted response elements is shown in orange, and the consensus sequence is colored green. The flanking sequence is red and shows CTE interactions and the triple basepairs between the inverted response elements. Nucleotides from the next asymmetric unit are boxed and the discontinuity in the phosphate backbone is denoted by a break symbol. Predicted hydrogen bonds are marked with black solid lines for major groove interactions and red for minor groove interactions. The van der Waals interactions occurring between T4 and Val589 are indicated by a dashed line. Water molecules are named W, and phosphate groups are dark teal. The number in superscript indicates nucleotide position within the response element. Protein DNA interface regions are indicated by the brackets under the model.

A, Ribbon diagram showing the intersubunit interactions of the PR dimer. Subunits are colored pink and green. Helix 1 (H1), helix 2 (H2), and helix 2′ (H2′) are indicated on each DBD subunit, with subunit A on the left. Proposed hydrogen bonds are marked with dashed red lines. Carbonyl and amine groups involved in potential hydrogen bonding are colored red and blue, respectively. B, Superpositioning of PR-DBD (blue) and ER-DBD (red) (35) structures shows the different positioning of Lys617 (K617). Amino acid numbering is for PR. Proposed hydrogen bonds are marked with dashed red lines and H1, H2, and H2′ are indicated.