Crystal structure of the human NKX2.5 homeodomain in complex with DNA target

Abstract

NKX2.5 is a homeodomain containing transcription factor regulating cardiac formation and function, and its mutations are linked to congenital heart disease. Here we provide the first report of the crystal structure of the NKX2.5 homeodomain in complex with double-stranded DNA of its endogenous target, locating within the proximal promoter -242 site of the atrial natriuretic factor gene. The crystal structure, determined at 1.8 Å resolution, demonstrates that NKX2.5 homeodomains occupy both DNA binding sites separated by five nucleotides without physical interaction between themselves. The two homeodomains show identical conformation despite the differences in the DNA sequences they bind, and no significant bending of the DNA was observed. Tyr54, absolutely conserved in NK2 family proteins, mediates sequence-specific interaction with the TAAG motif. This high resolution crystal structure of NKX2.5 protein provides a detailed picture of protein and DNA interactions, which allows us to predict DNA binding of mutants identified in human patients.

Figure 1

The schematic diagram of NKX2.5, NK family HD sequence alignment and ANF-242 DNA sequence. (A) A schematic model of NKX2.5. Both N-terminal (1–137) and C-terminal (198–324) domains contain A/P rich sequences and HD (138–197) is centrally located. (B) Sequence alignment of NK family HD domains. The numbering is based on the conventional HD numbers where NKX2.5 138–197 corresponds to HD residue 1–60. Secondary structure elements are represented using cyan cylinders for α-helices and black lines for coils on top of the sequences. Missense mutations identified in NKX2.5 HD were shown below the sequences. (C) ANF-242 DNA sequence used for the structural study. Palindromic NK2 elements are shaded in cyan.

Figure 2

Overall structure of NKX2.5 HD/ANF-242 complex. (A) Two different views are shown of the protein–DNA complex. The HD is shown in cyan, and the DNA is shown in stick representation with carbon in yellow, oxygen in red, nitrogen in blue, and phosphorus in orange. On the left diagram, both HDs are positioned on top of the DNA, and the view on the right was generated by rotating the molecule 90° along the horizontal axis. The helices and the N- and C-termini are labeled. (B) Side chains of the residues involved in formation of the hydrophobic core are shown in stick and van der Waals sphere representation. Only one HD is shown in this figure.

Figure 3

NKX2.5 HD and ANF-242 DNA interaction. (A) A diagram of HD-DNA contact in a half site of ANF-242 DNA. Hydrogen bonds and van der Waals interactions are represented with solid red and dotted blue lines, respectively. Numbered water molecules are shown in cyan boxes. (B) Structure of HD monomer and the DNA. Side chains of amino acid residues involved in DNA binding are shown with carbon in cyan, oxygen in red, and nitrogen in blue. The DNA model is shown in stick representation. Three helices and the N-terminus of the HD, and the major and minor grooves of DNA are labeled. (C–E) Detailed view of HD-DNA interactions through the N-terminus (C), the N-terminal section of α3 (D) and the C-terminal section of α3 of the HD (E). DNA model is shown with carbon in white using stick and surface representation. The HD is shown in stick and ribbon representation with carbon in cyan. For all molecules, oxygen is shown in red, nitrogen in blue, and phosphorus in orange. Hydrogen bonds are depicted using dotted black lines. Water molecules are shown as gray spheres. (F, G) Tyr54 and interacting DNA region. A stick model of Tyr54 side chain is shown with 2F_o – F_c map (1 σ level) calculated at 1.8 Å resolution (F) and with van der Waals surface representation (G).

Figure 4

Comparisons of wild type and HD mutants, N51K (A) and R53H (B), interactions. Energy minimized models of wild type, N51K, and R53H are shown in stick representation. Color scheme follows that of Figure 3 except in N51K and R53H mutants, carbon is shown in yellow. Dotted lines depict hydrogen bonds.