Computational Analysis and Molecular Design of Aromatic Stacking System Between Human ARTD2 Catalytic Domain and its Small-Molecule Inhibitors and Peptidic Ligands

Abstract

ADP-ribosyltransferase 2 (ARTD2) acts as a DNA break sensor and catalyzes the synthesis of polymers of ADP-ribose covalently attached to acceptor proteins at DNA damage sites. Its catalytic domain (CAT) contains a variety of aromatic amino aid residues that can form multiple [Formula: see text]-stacking interactions with substrates and inhibitors. Here, we systematically examined the intermolecular interaction of ARTD2 CAT domain with its cognate small-molecule inhibitors using hybrid quantum mechanics calculations and found that [Formula: see text]-stacking plays a critical role in the ARTD2-inhibitor recognition and association, where a number of ARTD2 inner aromatic residues contribute significantly to the inhibitor binding. A further molecular design of peptidic ligands was performed based on the structural basis and energetic property of the [Formula: see text]-stacking system involved in ARTD2/inhibitor interaction. These peptidic ligands are all [Formula: see text] electron-rich and composed of diverse unnatural aromatic amino acids. Interaction analysis revealed that the relative contribution of the [Formula: see text]-stacking system to ARTD2 affinity was improved from small-molecule inhibitors to peptidic ligands, although their total binding energies are roughly comparable. Subsequently, the computational findings were substantiated with biochemical assays, where several representative peptidic ligands were determined to have satisfactory inhibitory profiles against ARTD2 with biological activity at the nanomolar level. Structural analysis of ARTD2/peptidic ligand complexes observed a multiple [Formula: see text]-stacking networks at the computationally modeled complex interface of ARTD2 CAT domain with designed peptidic ligands, which can work with other nonbonded interactions such as hydrogen bonds and hydrophobic contacts to confer stability and selectivity for the complex recognition and association.