Developing a computational platform for induced-fit and chemogenetic drug design

PROJECT SUMMARY Prescription opioid therapy plays a critical role in the clinical management of pain in multiple acute and chronic settings. The challenges of effective pain management have led to over 2 million adults in the US, and over 12 million globally, with an opioid use disorder (OUD). OUD accounts for over 120,000 deaths annually worldwide. The dominant target of therapeutic opioids is the µ-opioid receptor (MOR). The analgesic effects of MOR agonists are due to Gα,i/o/z-protein signaling, and it has been proposed that undesirable side-effects of MOR agonists, such as respiratory depression and tolerance, can be mitigated through partial recruitment of Gi/o/z- protein subtypes. Thus, it is of clinical interest to determine the relationship between MOR signaling and analgesia versus side-effects to guide the design of therapeutic agonists that selectively activate the desired signaling pathway. G-protein coupled receptors (GPCRs), including MOR, are known to adopt a range of different functionally distinct configurations upon engaging orthosteric modulators and/or intracellular effector proteins. These induced-fit structural rearrangements cannot be modeled with existing computer-aided drug discovery algorithms during docking or design due to the time and resources required. It is the objective of this proposal to develop a customizable, multi-purpose computer-aided drug design (CADD) platform that can efficiently model largescale induced-fit conformational changes during small molecule and/or receptor sequence design. Completion of the proposal will enable structure- based design of biased agonists and DREADDs (Designer Receptors Exclusively Activated by Designer Drugs). This proposal will include innovative algorithms that leverage deep learning protein structure prediction methods and ultra-large make-on-demand chemical libraries to rapidly screen synthetically accessible molecules for those that can induce conformational changes required to activate G¬i¬-protein signaling in MOR. In collaboration, I will synthesize (Dr. Craig Lindsley), functionally validate (Drs. Craig Lindsley, Heidi Hamm, and Vsevolod Gurevich), and structurally characterize (Drs. Beili Wu and Matthias Elgeti) designed molecules and DREADDs. Experimentally validated partial and biased agonists and DREADDs will be fed back into the computational platform to be used as starting points for subsequent rounds of optimization. In this way, we will establish a computational-experimental iterative feedback loop.