Analysis of the Asymmetry of Activated EPO Receptor Enables Designing Small Molecule Agonists

Amgen solved the high-resolution cocrystal structure of erythropoietin (EPO) bound to the extracellular part of the receptor (EPOR) in 1998, which reveals that the EPO-EPOR interaction surface is formed by 11 salt bridges, 17 H-bonds, and 2 hydrophobic clusters centered at a pair of crucial phenylalanines (F93). The EPOR has two domains, one that penetrates the membrane and a second extracellular domain that forms one arm of the binding site for the EPO ligand. The complete competent receptor-binding site is a homodimer of EPOR with the two arms forming a funnel-shaped cup where EPO binds. The two binding arms of the EPOR dimer meet at the membrane at a 120 degree angle, which Amgen characterizes as, "erythropoietin imposes a unique angular relationship and orientation that is responsible for optimal signaling." They come to this conclusion, because the EPOR cocrystallized with 2 equivalents of a 20 residue EPO mimetic peptide created at Robert Wood Johnson (RWJ) activates the receptor with a 3 order of magnitude reduction in potency, and the binding arms are forced to meet at the membrane with an angle of 180 degrees. The vast interaction surface between EPO and EPOR forms a singularly important three-dimensional structure responsible for hematopoietic stem cell proliferation and differentiation-this is Amgen's conclusion. This goal of this work is to present experimental and computational evidence that the Amgen structure is a postsignaling off-state and that the RWJ structure with the partially active peptide mimetics is an on-state. A detailed side-by-side comparison of the two structures will be presented along with literature evidence that calls into question the Amgen claim that their structure is a unique on-state. A computational fragment-based drug discovery method applied to the RWJ structure was used to locate and characterize a new predicted small molecule binding site and a fragment analysis was performed based on theories of asymmetry to create a proposed agonist with MW<300. When this molecule was experimentally tested, it displaced radiolabeled EPO with nanomolar potency and transformed human hematopoietic stem cells into red blood cells with subnanomolar potency. Obviously, this small molecule makes none of the EPO-EPOR interactions that Amgen stated were essential for fully turning on the receptor and provides strong evidence that stabilizing receptor asymmetry, not specific interactions, is the critical factor needed for activating signal transduction. Finally, when the agonist was altered to remove the asymmetric component, it still was able to displace radiolabeled EPO in competition binding experiments, but it no longer activated the receptor.