Distinguishing Different Hydrogen-Bonded Helices in Proteins by Efficient ¹H‑Detected Three-Dimensional Solid-State NMR

Helical structures in proteins include not only α-helices but also 3₁₀ and π helices. These secondary structures differ in the registry of the CO···H–N hydrogen bonds, which are i to i + 4 for α-helices, i to i + 3 for 3₁₀ helices, and i to i + 5 for π-helices. The standard NMR observable of protein secondary structures are chemical shifts, which are, however, insensitive to the precise type of helices. Here, we introduce a three-dimensional (3D) ¹H-detected experiment that measures and assigns CO–H^N cross-peaks to distinguish the different types of hydrogen-bonded helices. This hCOhNH experiment combines efficient cross-polarization from CO to H^N with ¹³C, ¹⁵N, and ¹H chemical shift correlation to detect the relative proximities of the CO_i–H_i+j^N spin pairs. We demonstrate this experiment on the membrane-bound transmembrane domain of the SARS-CoV-2 envelope (E) protein (ETM). We show that the C-terminal five residues of ETM form a 3₁₀-helix, whereas the rest of the transmembrane domain have CO_i–H_i+4^N hydrogen bonds that are characteristic of α-helices. This result confirms the recent high-resolution solid-state NMR structure of the open state of ETM, which was solved in the absence of explicit hydrogen-bonding restraints. This C-terminal 3₁₀ helix may facilitate proton and calcium conduction across the hydrophobic gate of the channel. This hCOhNH experiment is generally applicable and can be used to distinguish not only different types of helices but also different types of β-strands and other hydrogen-bonded conformations in proteins.