Abstract #5352: Coumarin based NQO1 inhibitors. Computational modelling, biochemical effects and modulation of protein-protein interactions

AACR Annual Meeting-- Apr 18-22, 2009; Denver, CO NQO1 is a homodimeric enzyme with one molecule of non-covalently bound FAD per monomer using either NADH or NADPH as reducing cofactors. The catalytic mechanism of NQO1 relies on sequential two electron transfer from NAD(P)H to FAD to the substrate. This \#8220;ping pong\#8221; reaction is inhibited competitively by dicoumarol. NQO1 is known to interact with other proteins in an NAD(P)H dependent manner and the interaction of NQO1 and p53 leads to stabilization of p53. Coumarin inhibitors of NQO1 block the catalytic activity of NQO1 and modulate the protein-protein interactions of NQO1 but the prototype inhibitor dicoumarol has multiple off target effects. In this work we describe the development of more specific coumarin-based NQO1 inhibitors that retain their ability to modify NQO1 protein-protein interactions. Lead compounds, including dicoumarol, were subject to molecular dynamic simulations of their interaction with NQO1 in aqueous solution. These studies indicated stable stacking interactions within the active site and revealed the presence of flexible regions in the protein. This strengthened our hypothesis that the enzyme inhibitors may have differential effects on p53 stability depending upon their ability to perturb the tertiary structure of NQO1. We have used a variety of biochemical techniques to evaluate the nature of binding of these inhibitors to NQO1. Isothermal calorimetry showed the compounds to have one of two binding modes such that some inhibitors bound to two identical active sites with equal affinity (one site per monomer) whereas others displayed positive cooperativity by acting as allosteric effectors. These findings were confirmed by the use of fluorescence spectroscopy, looking specifically at the changes in environment of tryptophan residues upon binding of inhibitors to NQO1. This technique also suggested two distinct binding modes. Computational studies have provided a means to evaluate the changes in protein dynamics and identify individual amino acids which may be involved in long-range interactions in different parts of the enzyme. Molecular dynamical simulations have allowed the comparison of covariance matrix maps from ligand bound to unbound protein and have revealed the presence of residue-residue correlated motions which change upon ligand binding. The computational studies correlate with the biochemical data and provide a basis for predicting protein-protein interactions. In conclusion, we have clear evidence from both computational and biochemical studies to show that analogues of dicoumarol can have profoundly different effects on the tertiary structure of NQO1. Further, we have preliminary evidence to show that different inhibitors have a differential effect on the ability of NQO1 to interact with its client proteins. Citation Information: In: Proc Am Assoc Cancer Res; 2009 Apr 18-22; Denver, CO. Philadelphia (PA): AACR; 2009. Abstract nr 5352.