Questions: NMR for Proteins

Question 1

Why is protein NMR typically limited to molecules below ~40 kDa, while X-ray crystallography and cryo-EM have no practical upper size limit?

• A. NMR magnets cannot produce strong enough fields for larger molecules
• B. Larger proteins tumble more slowly in solution, which increases transverse relaxation rates (broader NMR lines) and causes severe spectral overlap — the signals from thousands of atoms become too broad and too overlapping to resolve
• C. Larger proteins are always insoluble, so NMR cannot be performed
• D. NMR requires crystallization, which fails for large proteins

Answer: (B)

NMR resolution depends on the linewidth of each resonance, which is inversely proportional to the transverse relaxation time (T2). T2 decreases as molecular tumbling slows (larger molecules tumble more slowly). Broader lines mean more spectral overlap, and with thousands of 1H, 13C, and 15N resonances in a protein, the spectrum becomes uninterpretable above ~40 kDa with standard methods. TROSY (Transverse Relaxation-Optimized Spectroscopy) selects for the slowest-relaxing component of each multiplet, extending the practical limit to ~100 kDa, but this requires deuteration and is still far more limited than crystallography or cryo-EM in terms of molecular size.

Question 2

NMR provides a single, unique structure for a protein, just like X-ray crystallography.

• T. True
• F. False

Answer: False

NMR structure determination produces an ensemble of structures, not a single one. The experimental data (NOE distance restraints, dihedral angle restraints, residual dipolar couplings) constrain the structure but do not uniquely determine it — there are typically more degrees of freedom than restraints. Simulated annealing or molecular dynamics calculations generate an ensemble of structures that all satisfy the experimental restraints equally well. This ensemble reflects both the uncertainty in the data and the genuine conformational dynamics of the protein in solution. Regions that are well-defined across the ensemble (low RMSD) are conformationally rigid; regions that vary (high RMSD) are genuinely flexible. This is actually an advantage — the ensemble captures biological dynamics that a single crystal structure cannot.

Question 3

What information does NMR provide about protein dynamics that crystallography cannot?

Model answer: NMR measures molecular motions over a wide range of timescales: backbone and side chain dynamics on the ps-ns timescale (from 15N and 13C relaxation measurements — order parameters quantifying the amplitude of motion), conformational exchange on the us-ms timescale (from relaxation dispersion experiments — detecting interconversion between distinct conformational states), and slow exchange on the ms-s timescale (from hydrogen-deuterium exchange — measuring solvent accessibility reflecting protein breathing motions). This multi-timescale dynamic information reveals which regions are rigid vs. flexible, which residues interconvert between multiple conformations relevant to function (enzyme catalysis, allosteric regulation, ligand binding), and which regions are transiently exposed. X-ray crystallography provides only B-factors (which conflate dynamics with crystal disorder) and cannot access these timescale-specific dynamic measurements.

NMR dynamics studies have revealed that enzyme catalysis often involves conformational fluctuations on the same timescale as the catalytic rate — suggesting that dynamics are rate-limiting. Allosteric communication has been shown to propagate through networks of dynamically coupled residues. These insights are invisible to static structural methods.