Accidental interaction between PDZ domains and diclofenac revealed by NMR-assisted virtual screening

Abstract

In silico approaches have become indispensable for drug discovery as well as drug repositioning and adverse effect prediction. We have developed the eF-seek program to predict protein-ligand interactions based on the surface structure of proteins using a clique search algorithm. We have also developed a special protein structure prediction pipeline and accumulated predicted 3D models in the Structural Atlas of the Human Genome (SAHG) database. Using this database, genome-wide prediction of non-peptide ligands for proteins in the human genome was performed, and a subset of predicted interactions including 14 PDZ domains was then confirmed by NMR titration. Surprisingly, diclofenac, a non-steroidal anti-inflammatory drug, was found to be a non-peptide PDZ domain ligand, which bound to 5 of 15 tested PDZ domains. The critical residues for the PDZ-diclofenac interaction were also determined. Pharmacological implications of the accidental PDZ-diclofenac interaction are further discussed.

Figure 1

Examples of NMR-binding assay between PDZ domains and the compound cocktails. (a) Each overlaid spectrum was derived from a PDZ domain with (red) and without (black) cocktail. Upper spectra show that no signal changes were observed with mixing of the PDZ domains and a cocktail. Lower spectra show the signal changes when PDZ domains were mixed with a cocktail. (b) Summary table of binding assays using compound cocktails. The number of the plus signs indicates the degree of signal changes (+, less than 15 signals changed; ++, less than 30; +++, over 30, respectively). The minus sign indicates no signal changes. PDZ1, PDZ3, and PDZ14 were not examined.

Figure 2

Examples of NMR titration of PDZ domains with each compound. Each overlaid spectrum was derived from a PDZ domain with (red) and without (black) the compound. Upper spectra show results where no signal changes were observed after mixing the PDZ domain with a compound. Lower spectra show the signal changes of the PDZ domain when mixed with a compound.

Figure 3

Identification of the interface between the mZO-1 PDZ1 domain and a compound. Mapping of residue signal changes upon mixing with DIF (a), FLF (b), and FUA (c) onto the ribbon model of the mouse ZO-1 PDZ1 domain (PDB:2RRM). (d) The ribbon model represents the canonical binding pocket between the PDZ domain and peptide (PDB:2H2B).

Figure 4

Comparison of the interfacial residues. Schematic of the interfacial residues of mZO-1 PDZ1 (a), PDZ13 (b), PDZ7 (c), PDZ8 (d), PDZ9 (e), and PDZ11 (f) are depicted. The canonical binding pocket lies between a β-sheet composed of β2 and β3 strands and α1 helix, represented as block arrows (β2 and β3 strands) and a cylinder, respectively. Each ellipse represents the location of each residue. The atypical residues of PDZ11 are in red.

Figure 5

Identification of the key residues of the mZO-1 PDZ1 domain responsible for DIF binding. Overlaid spectra derived from wild type (a) and the V92E-R96Q mutant (b) of the mZO-1 PDZ1 domain with (red) and without (black) DIF.