SUMMARY The proposed research focuses on processes that are mediated by proteins, that are crucial for cellular physiology and that are dysregulated in human disease. This effort has two components that differ topically but share common themes in experimental design. Fundamental mechanistic insight is our primary goal, but outcomes could support future therapeutic advances. We use chemically synthesized polypeptides to test molecular hypotheses, an approach that transcends the compositional limitations of biosynthetic polypeptides. One major focus involves signal transduction by a set of G protein-coupled receptors (GPCRs) that are naturally modulated by long polypeptide hormones (B family GPCRs). Agents that activate these receptors are used in human medicine. We conduct parallel studies with multiple receptors so that we can discern which features of the signal transduction process are conserved and which features are unique to specific receptors. We have previously explored the parathyroid hormone receptor-1 (PTHR1) and the glucagon-like peptide-1 receptor (GLP-1R), agonists of which are used to treat osteoporosis or type 2 diabetes, respectively. Continuing work on these two receptors will be complemented by new studies with the glucagon receptor (GCGR; glucagon is used by diabetes patients to reverse hypoglycemic shock) and the GLP-2 receptor (GLP-2R; agonists are used to treat short-bowel syndrome). Proposed studies will harness unique capabilities of our group to explore the consequences of varying agonist conformational propensities in terms of signaling outcomes and receptor binding. Results should be very impactful because agonist design efforts motivated by therapeutic goals usually focus on stabilizing a specific receptor-bound conformation of the agonist, which is fully α-helical for the GPCR agonists we study. Our recent findings raise questions about this approach; we have discovered surprisingly potent agonists based on designs that destabilize the α-helical conformation. The other major research focus is liquid-liquid phase separation (LLPS) mediated by proteins. LLPS in cells leads to a variety of “membraneless organelles” (MOs) that appear to be crucial for cellular function. The noncovalent interactions that underlie these dynamic assemblies are not well understood. Evolution of dynamic MOs toward more ordered assemblies appears to underlie some neurodegenerative diseases. Our approach involves developing models of protein-mediated LLPS in which one component is a peptide of < 50 residues and therefore readily accessible via chemical synthesis; the other component is a larger biopolymer (protein or RNA). This design strategy enables incisive evaluation of the roles of diverse noncovalent interactions in the dynamic associations that are unique to MOs. Results should inform future efforts to address diseases associated with “hardening” of MOs within cells.