NOESY and Distance Constraints

Core Idea

The Nuclear Overhauser Effect Spectroscopy (NOESY) experiment detects through-space proximity between hydrogen atoms by measuring the cross-relaxation between nuclei that are close in three-dimensional space (typically less than 5 Angstroms), regardless of their connectivity through covalent bonds. NOESY cross-peaks provide distance restraints — the closer two protons, the stronger the NOE signal. A network of thousands of such distance restraints (short, medium, and long-range) provides the primary experimental data for NMR protein structure determination. Long-range NOEs (between residues far apart in sequence but close in space) are the most valuable because they define the protein's three-dimensional fold.

Explainer

The Nuclear Overhauser Effect is the physical phenomenon that makes NMR protein structure determination possible. When two hydrogen atoms are close in space, their nuclear spins interact through a process called cross-relaxation: if one spin is perturbed (by radiofrequency irradiation), it affects the magnetization of its nearby neighbors. This interaction depends on the distance between the nuclei — specifically, the cross-relaxation rate is proportional to 1/r⁶ — making it extraordinarily sensitive to proximity. The NOESY experiment measures these cross-relaxation interactions systematically, producing a 2D spectrum where each cross-peak connects two protons that are close in space.

The key insight is that the NOE is a through-space interaction — it reports on three-dimensional proximity regardless of covalent connectivity. Two protons can be 100 residues apart in the amino acid sequence but produce a strong NOE if they are less than 5 Angstroms apart in the folded protein. This is exactly the information needed to determine the 3D fold: NOEs between sequentially distant but spatially close residues reveal how the polypeptide chain folds back on itself, how helices pack against sheets, and how the hydrophobic core is organized.

Structure determination from NOE data is a constraint satisfaction problem. Each observed NOE provides an upper-bound distance restraint: the two protons must be within ~5 Angstroms (for a weak NOE) or within ~2.5 Angstroms (for a strong NOE). A typical well-determined NMR structure uses 2,000-4,000 NOE distance restraints, supplemented by backbone dihedral angle restraints (from chemical shifts via TALOS) and sometimes residual dipolar couplings (from partial molecular alignment). Computational algorithms (simulated annealing in torsion angle space, implemented in programs like CYANA and Xplor-NIH) search for structures that simultaneously satisfy all restraints while maintaining good stereochemistry. The result is an ensemble of 20-40 structures, all consistent with the data, whose convergence (or lack thereof) directly reveals which regions are well-defined and which are flexible.

The practical challenges include spectral overlap (many protons have similar chemical shifts, making it hard to identify which peaks are which), spin diffusion (NOE transfer through intermediate protons can generate artifactual long-range NOEs), and dynamics (conformational averaging can modulate NOE intensities). Three-dimensional and four-dimensional NMR experiments (separating protons by their attached 13C or 15N chemical shift) address overlap, and careful analysis protocols handle spin diffusion and dynamics. Despite these challenges, NMR structure determination by NOESY distance restraints has produced thousands of protein structures in the PDB, uniquely capturing the solution-state, dynamic nature of biomolecules.