There is always a new therapy to discovered, either by overcoming side effects or other liabilities in current drugs, or through exploring new technologies, proteins, or diseases. Focusing on small molecules, the initial stages of the drug discovery process require the identification of molecules modulating a protein of interest. In Chapter 1, one approach available for discovery is described – molecular docking to discover new small molecules of interest from virtual chemical databases. This thesis applies the approach to overcome side effects of current opioid pain drugs, to identify new antivirals for SARS-CoV-2, and to probe larger applications of the growing chemical space.
Chapter 2 describes how the computational approach identifies new nonopioid pain therapeutics acting through the α2A-adrenergic receptor (α2AAR). Here, we show the translational impact of computational predictions as many are efficacious in mice. We argue a case for further drug discovery focused on this protein, because not only were the opioid side effects avoided, but also a characteristic α2AAR-caused side effect is not seen with new compounds.
Chapter 3 demonstrates application of the computational technique to an understudied viral protein, the SARS-CoV-2 nonstructural protein 14 (nsp14), using different subsets of the chemical databases for targeted results. Here, ‘traditional’ noncovalent ‘lead-like’ subsets, smaller ‘fragment’ molecules, and covalent molecules are utilized and successfully identify nsp14 inhibitors. We show in a pandemic-driven project that multiple techniques are needed to progress towards antivirals inhibiting an understudied protein.
Chapter 4 continues the themes of Chapter 3, but with the SARS-CoV-2 main protease (MPro). Here, we use a similar approach with different types of chemistry found in the databases and identify numerous inhibitors. Structural studies demonstrate the computationally predicted and experimental ligand geometries are in high agreement. Finally, we begin to probe the antiviral efficacy of these new compounds and contribute the knowledge to the field for further antiviral development.
Chapter 5 focuses on a new application of the virtual screening of chemical databases. Here, we ask if this approach can be applied prospectively to discover molecules with polypharmacology – having designed activities on multiple proteins of interest. We simulate the effect of the growing chemical databases to understand their effect on docking for polypharmacology. This chapter also blends into the most immediate future directions of the thesis work and is described in context of pain therapeutic discovery.