Hydrophobic Amino Acids as Universal Elements of Protein-Induced DNA Structure Deformation

Abstract

Interaction with the DNA minor groove is a significant contributor to specific sequence recognition in selected families of DNA-binding proteins. Based on a statistical analysis of 3D structures of protein-DNA complexes, we propose that distortion of the DNA minor groove resulting from interactions with hydrophobic amino acid residues is a universal element of protein-DNA recognition. We provide evidence to support this by associating each DNA minor groove-binding amino acid residue with the local dimensions of the DNA double helix using a novel algorithm. The widened DNA minor grooves are associated with high GC content. However, some AT-rich sequences contacted by hydrophobic amino acids (e.g., phenylalanine) display extreme values of minor groove width as well. For a number of hydrophobic amino acids, distinct secondary structure preferences could be identified for residues interacting with the widened DNA minor groove. These results hold even after discarding the most populous families of minor groove-binding proteins.

Keywords: DNA shape; hydrophobic; indirect readout; minor groove; protein–DNA interaction; specific recognition.

Figure 1

An overview of the workflow and analyses performed in this work. Individual steps are described in detail in Section 3.

Figure 2

Frequencies of amino acids contacting narrow, standard, and wide DNA minor grooves (DS1 dataset).

Figure 3

Frequencies of amino acids contacting narrow, standard, and wide DNA minor grooves (DS2 dataset).

Figure 4

Distributions of DNA minor groove widths associated with the dinucleotides bound by the respective amino acids (top: DS1, bottom: DS2). Amino acids are sorted in descending order according to the associated median minor groove width. Whiskers of the boxplots denote the first and last 25% of the distributions. Maximal length of a whisker was set to $1.5$ times the interquartile range. Any observations belonging to the whisker range that is further from the first/third quartile than this threshold is denoted as an outlier and marked as a dot in the plots.

Figure 5

Distributions of DNA minor groove widths corresponding to hexamers with different values of GC content (complete non-redundant dataset). The numbers of examples refer to the number of hexamers with said GC content.

Figure 6

TBP DNA-binding motif interacting with the extremely widened minor groove. The contacted DNA sequence consists only of AT base pairs (green). Two phenylalanine residues (orange) intercalate into the helix on one or both edges of the motif (detail on the right). A similar DNA-contacting motif appears in the dataset also in 4ROC and 1YTB structures, both containing similar extreme deformations of the minor groove in AT-rich sequences.

Figure 7

The non-intercalating interaction in the structure 5JH0. Bulky, hydrophobic residues (Phe, Ile) are shown in teal. Visualized using PyMOL-1.7.2.1 [27].

Figure 8

Positioning of glutamic acid residues with regards to the DNA dinucleotide. Residue–dinucleotide steps were superimposed using the Kabsch algorithm. All dinucleotide steps are represented by the step with minimal RMSD from all other steps.

Figure 9

Illustration of the minor groove width categories: (a) Narrow minor groove (T4 Endonuclease VII H43N mutant in complex with heteroduplex DNA, PDB ID 2QNF); (b) standard minor groove (human Pax-6 paired domain–DNA complex, PDB ID 6PAX); (c) wide minor groove, (topoisomerase II alpha bound to DNA, PDB ID 4FM9). In this case, the wide minor groove is associated with bending of the DNA structure. The minor groove nitrogen and oxygen atoms are colored blue and red, respectively.