Development of Quantitative Solid-State NMR Methods to Characterize Membrane Proteins

Abstract

Membrane proteins are critical components of all cells and viruses. While about one quarter of human proteins are membrane proteins, they constitute half of all drug targets. Despite their importance, membrane proteins are underrepresented among known protein structures, constituting only about 2 percent of the Protein Data Bank. This discrepancy is due to the unique difficulties in studying membrane proteins, which make many techniques commonly used in structural biology extremely challenging. Membrane proteins often display a structural dependence on the local environment. It is therefore essential to have structural biology tools to study these critical proteins in native-like environments. Solid-state Nuclear Magnetic Resonance (NMR) spectroscopy provides one of the few methods available to study the structure and dynamics of membrane proteins directly in the lipid bilayer. Herein, practical and theoretical considerations of dipolar and chemical shift anisotropy recoupling experiments are presented. These experiments were applied to the study of membrane proteins. New experiments and novel analysis techniques were developed, and the results guided biophysical understanding and drug development.

Among the membrane proteins of SARS-CoV-2, the Envelope (E) protein is the least understood. E forms a membrane-bound ion channel and is associated with inducing the respiratory symptoms of the disease. The exact oligomeric state of E was not known. The fluorine centerband-only detection of exchange (CODEX) experiment was employed to directly measure the oligomeric state of E in lipid bilayers. The transmembrane domain of E forms a pentamer, while a construct including the ectodomain forms a dimer. Under certain conditions, the pentamers cluster together, forming supramolecular assemblies that may have a unique role in the virus life cycle.

New sensitivity-enhanced carbon-fluorine rotational-echo double-resonance (REDOR) experiments are developed and used to investigate the drug binding of E. The small molecule drug hexamethylene amiloride binds to E at the protein-lipid interface. This informed the development of higher affinity inhibitors, which were also shown to bind E at the lipid interface. A novel strategy to identify ligand binding sites of proteins without sequential resonance assignment is presented. The technique uses a computationally efficient second moment approximation to calculate REDOR dephasing, and simulated annealing to explore the associated parameter space.

The new methods and advances in quantification and simulation of the REDOR and CODEX experiments enhance the available solid-state NMR toolkit for the study of critical membrane proteins.