Identification of small molecules targeting the MdmX-p53 Interaction in Human Cancers
In about 50% of human cancers, the p53 gene is mutated while in the remaining cancers, the p53 pathway is dysregulated. One mechanism is through overexpression of negative regulators of p53 such as MDMX and MDM2. These two proteins bind to p53 and inhibit its function by directing p53 for degradation. The N-terminal domains of MDMX and MDM2 that bind to p53 are highly similar in sequence and in structure. Despite these similarities, the binding of small molecules in their p53 binding pockets are very distinct. For example, nutlin-3a, an inhibitor of the MDMX/2-p53 binding interaction, binds more than 200-fold weaker to MDMX than MDM2. In fact, there is currently no MDMX inhibitor in clinical trials. Inhibition of MDM2 alone is insufficient in some cases, particularly if MDMX is overexpressed such as in retinoblastoma. Additionally, there is evidence that MDM2 and MDMX have distinct and non-redundant roles. The purpose of this project is to find molecules that can bind to MDMX and compete with p53. To achieve this goal, we turn to computational methods to find small molecules that can target the p53 binding site of MDMX in collaboration with Atomwise through an Artificial Intelligence Molecular Screen (AIMS) award. Using their proprietary neural network-based protocols, Atomwise targeted the p53 binding site of MDMX which provided us with seventy-two molecules. We employed Fluorescence Polarization assay to screen these compounds for their activity against MDMX. Using this preliminary assay, we have identified seven possible molecules. Using NMR, we showed that at least one of these compounds bind to MDMX. We are currently in the process of validating the other molecules. In the future, we will perform dose curves to determine the EC50 of the compounds in vitro as well as isothermal titration calorimetry to determine their binding affinity to MDMX. Importantly, we will also screen against MDM2 to determine differential affinity against MDMX and MDM2. If successful, we will proceed to human cell cultures in cancer models that show overexpressed MDMX.